Local Legend 10: Boeing 307 Stratoliner

Ça, c’est bon, ça! Tres bien! Merci beaucoup! :beers: And you can tell it’s Mardi Gras time down here on the bayou or I wouldn’t be drunk enough to talk like that :wink:

1 Like

I have managed to get a custom engines.cfg file to override the encrypted version, which means I now have control of the engines. I’m in the process of learning about and tuning the significant number of available variables.

One of the aspects of this project that has been extremely interesting to me is the deep dive I’m taking into the Wright R-1820 Cyclone engine. I love history, I love technical documentation, I love research, I love reading and I love vintage aviation, so this is a boon.

To start with, there is a massive number of R-1820 variants. I’ve scored a fair amount of original Wright R-1820 documentation in .pdf format, including an original Wright production spreadsheet breaking down some key information on each of these variants.

It’s a bit of a weird chart in that Wright grouped a number of engine variants in a section and then used a large brace (aka curly bracket) to list the aircraft these were fitted to. In other words, it is imprecise in that it lists a number of engines and a number of aircraft and lumps them together.

Wikipedia is not known for being 100% accurate, but it has two engines listed for the 307, which correlate with the imprecise Wright chart:

  • 307: Wright GR-1820-G102A Cyclone
  • 307B: Wright GR-1820-G105A Cyclone

Both engines output 1100 HP, but the G105A has a two-speed supercharger for higher altitude performance.

In addition to this, the “Airplane Flight Manual” for 307 serial number 2003 has a print date of March 23, 1954 — well past the end of WWII. In this manual the engines are listed as GR-1820-97. That variant of the Cyclone has the largest (huge) production number and, unsurprisingly, was fitted to the B-17. #2003 was owned by Pan Am and did not see wartime service by the USAAF. However, by 1954 #2003 would have undergone significant engine work and available surplus R-1820 engines would have, overwhelmingly, been those B-17 variants.

The B-17, though, was turbo-supercharged (single-speed mechanical supercharger and exhaust-driven turbocharger) and I believe the turbos were not fitted to the 307 for two reasons:

  1. The engine nacelles and cowlings would have needed to be modified to accommodate the extra mechanicals and ducting for the turbochargers and intercoolers. Additionally, there would have to be controls fitted in the cockpit for controlling the turbo-superchargers and the aircraft’s standard operating procedures would have been quite different to that of the original specification. Reading a B-17 pilot training manual illustrates just how significant these differences are.
  2. The engine performance charts listed in the 1954 “Airplane Flight Manual” have lower performance characteristics than those could I find on the B-17

Lastly, when the USAAF returned the five conscripted TWA 307Bs – now designated C-75 – to TWA after WWII, these were refitted with 1200 HP GR-1820-G205A Cyclone engines with two-speed superchargers.

In short, I have four distinct engines with four different performance characteristics, but the only performance charts for the 307 I can find are those for #2003 as fitted with B-17 engines sans turbos in 1954. #2003 is the only surviving (and airworthy) 307, which sits in the Smithsonian and is highly likely the aircraft our 307 is modeled after.

As I make progress on this, I plan to include two different engine profiles for the mod – one for single-speed and one for two-speed superchargers. Since we don’t have a supercharger shift lever in the 307, nor does the SDK offer a way to change gears manually, this will be an automatic function that occurs at critical altitude for the first speed. Of course, I’m going to have to use theoretical values to come up with performance characteristics, but that is nothing new when it comes to these simulated vintage aircraft.

Taking all of this information and then translating this into the SDK’s engine variables is extremely fascinating.


I’ve heard two things on the superchargers. On the one hand, Flying Iron (my absolute favorite developer) says they’re not modeled in game and therefore they made the Hellcat’s supercharger automatic saying it’d revert to manual if the game allowed for it later on. In case you don’t own it, basically it kicks in at 12000 and again at 26000 feet.

On the other hand, PMDG modeled a supercharger for their DC-6. At the proper altitude you throw the switches and voila, more power.

Probably beside the point since, as you say, we don’t have a lever in the 307.

I’m sure you’re right that this was modeled after that lone surviving plane.

Superchargers are definitely modeled in-game. In fact, they have a legacy mode and a new mode with more customization options – mainly up to 5 supercharger gears.

What isn’t modeled is the ability to shift the supercharger’s gears via manual control, hence Flying Iron using the automatic method employed by the SDK.

PMDG definitely had to create something custom for this. I’ve also examined the Duckworks mod for the DC-3 and that has some custom trickery, too.


There are also 3rd and 4th reasons. #3: There are many pics of the undersides of 307 nacelles and they quite obviously don’t have the turbo exhaust pipes of the B-17. #4: Also, pre-WW2, the US regarded its turbos as a military secret so they weren’t available for civilian use until after the war IIRC, even to companies (like Boeing) building both military and civilian planes.

As to superchargers, by the 1930s, EVERY Pratt or Wright radial had at least a 1-stage centrifugal supercharger as an integral part of the engine because this was just the perfect fit. It slung the fuel-air mix evenly to all the radial cylinders so basically functioned as an engine-driven fuel injector, but it had the beneficial side-effect of being able to cram more air into to the mix than the baro allowed for naturally aspirated engines, thus increasing performance above natural aspiration from say 3000’ up to about 16000’ or so. Above that, to maintain approximately the power you had at lower altitudes, you needed either a 2nd stage or gear for the supercharger or a turbo.

However, apparently 2-stage superchargers weren’t at all common (outside of military applications, and even rarely then) before WW2, although they became so during the war. And once that happened, EVERY radial then had 2-stage superchargers, just like they’d previously all had 1-stage.

As to coding for this sort of thing, the Duckworks mod for the 40th Anniversary DC-3 has made the 2nd stage available with levers to pull. The stock plane has the levers but they are inop. Also, the PMDG DC-6 and the Red Wing Connie both have 2nd stages with controls to shift gears. So making it work in the code is definitely doable. But if the 307 has no levers for it, it’s probably safe to assume that it didn’t have 2-stage superchargers until refitted with newer engines post-war.


the Duckworks mod for the 40th Anniversary DC-3 has made the 2nd stage available with levers to pull

This is legitimately one of my favorite things about the Duckworks mod. What a game changer for the DC-3.

Engine performance tuning update:

I’m still, very much, in the throes of tuning values, but so far I’ve got a much more powerful aircraft that can climb on METO power (MP 39.5" / RPM 2300) to 12500’ at 1000’/min (my conservative climb speed) with ease. It then shifts into second stage supercharging and continues to climb up to 20000’ where MP closes in on critical for the second stage at MP 31" firewalled.

I’m currently leveled off at 20000’ and cruising with MP 27" / 2100 RPM at 202 MPH ground speed.

We, test pilots, are all having to wear oxygen masks, because the cabin pressurization system is non-op, but that’s a fair trade-off to be up here in the rarified air.

I, suddenly, like this aircraft a whole lot more…


Oh, and I’ve enabled mixture control. It currently requires manual adjustment, but I can, at least, control fuel flow now.


Music to my ears!!

You know we had to take all that pressurization stuff away to carry some extra troopers.
But now you have it back so don’t cry.

1 Like

Will you be able to make the fuel pumps functional?

Without revealing my observations regarding the pumps, could you tell me what you’re experiencing?

Absolutely nothing. The pressure doesn’t change when you turn them on or off. Starting the aircraft isn’t affected (that I can see) by turning them on of off. It’s not a big deal to me, just something that’s “wrong” with the plane :slight_smile:


In the “stock” un-modded 307 here’s what I experience: With engines off and fuel tank selectors set to how they are set on C&D (Inboard Extra), turning on a booster pump will indicate 0 PSI on the Fuel Pressure gauge. If you rotate a pump’s matching tank selector valve to Inboard Extra, the pressure will jump to 11 PSI and remain there even if you switch the valve back to Onboard Extra – the pressure will still show 11 PSI. There is some glitch with the valves and the pumps when spawning.

With my published flightsim.to mod you won’t see this behavior since you have to set the tank selector valves from OFF to a tank of choice from the outset.

The other thing that is a little strange is that the “Boost” switches on the starter switch panel are repeats of the four booster switches at the bottom of the same panel, so that is totally odd. I can’t find any good info on those switches or the other four booster pump switches. I see them in photos of the Smithsonian 307, but I don’t have a high enough resolution photo to zoom in and read the labels. I don’t see them in early B&W cockpit photos of, possibly, other 307s.

On the DC-6 and the Twin Beech (a real one, not Carenado’s) the Boost switches that are next to the Starter switches are ignition boosters not fuel boosters.

Nothing of the sort is mentioned in the, rather sparce, 307 Flight Manual I have. It mentions Booster pumps, but not ignition boosters nor does it indicate where these switches are located. There are no illustrations nor photos, either.

Incidentally, the engines won’t start if you don’t turn on one of those booster pump switches (either the one paired with the starter and primer switch or the separate one with the on/off toggle switch) when starting the aircraft with my mod installed.

1 Like

Yea, I think I was just using the ones by the starter switches. Fuel pressure seems extremely low. You have to keep the RPM up very high to avoid the yellow warning lights. I was just assuming the fuel boost pumps (or w/e they are) would alleviate that situation.

The fuel pressure modeling is, indeed, not right.

Unfortunately, the SDK’s complex fuel system is in use rather than the simple fuel system.

This means the parameters for the fuel system are embedded in the encrypted flight_model.cfg file and the fuel system values in engines.cfg are for the simple system.

In the interim, we’re going to have to live with the poorly modeled fuel pressures, until I go down the flight_model.cfg route…

1 Like

The March of 1954 one ? This book should be used with caution … For example :

The R1820-97 is as powerful at 39.5" / 2300 rpm than at 43" / 2400 rpm. Riiiiiiight.

As for the switches near the engine starters, it would indeed make more sense for them to be ignition boosts.

Many thanks for your efforts to save the 307 from Hangarqueenship though !


To add to the many annoying things of the MSFS 307 … Propeller feathering has no visual effect on the 3D model. Not even sure it has an effect on aero drag either in fact … The engine was certainly in no hurry to stop, even at 100 kts IAS after having been shutdown for a while.

This plane is what you car restoration guys call a “project” I think. :sweat_smile:

1 Like

I certainly thought that odd, too.

When feathering on the ground, I do see the blades instantly flip to some sort of visual feathered position (no animation present). More importantly, I definitely have control over the propeller’s characteristics, including beta pitch range and feather position. In addition, I have control over the internal friction for the engine both whilst running and for shutdown.

Again, I’m still really into the early part of refining what I’ve done here, so I haven’t yet attempted an in-flight engine shutdown and feather, but I’ll get to it and should be able to get them to behave in a realistic fashion.

I want to feel that way, too, but looking at the overhead panel in old B&W images, I don’t see any other cluster of 4 switches that would be akin to the 4 bespoke on/off booster switches we have with our 307. I do see those four switches in images of the restored aircraft, though, but can’t read the labels. I wonder if they are oil dilution switches rather than boosters. Basically, the early aircraft had to have booster pump switches and the only place that I can, easily, see them being would be on that start/boost/prime switch cluster.

I was hoping the B-17 had a similar arrangement, but it doesn’t. The fuel boosters are on a separate panel from the start switches. There are also no ignition booster switches.

For sure and it’s a real “labor of love” kind of approach to fixing it, too.

The SDK is pretty vague on how many of its variables’ values will translate into what you see in terms of performance changes in-sim. I’m just having to change a value in a radical way, reload the sim and see what the change did. It’s a fairly slow process. I’ve totally stripped the sim down to the, true, absolute base-sim and just the 307. It loads much more quickly, fortunately.

This whole process reminds me much of tuning a race car — both drivetrain-wise and suspension-wise. You make a single massive change to learn its effect, then bring it back to a baseline and work on incremental changes to then refine it from there.

I feel pretty encouraged, though. The one flight I took yesterday was the first time I’d ever flown the 307 where I honestly enjoyed it and didn’t end up upside down in a pile of imaginary wreckage. The take-off, climb, cruise, descent and landing felt much more like the way the DC-6 behaves. I felt I had so much more control over the aircraft in a more consistent way and this was all due to being, unsurprisingly, able to maintain airspeed and having the ability to apply power and see change.

What I find so odd about its default characteristics is how much it struggles. Its baseline C&D payload is 50% fuel, the two pilots and a flight engineer at 170 lbs each. It’s essentially empty and, yet, It performs so poorly.

1 Like

But you don’t have to reload the sim ? Enter Dev mode, create a project (doesn’t matter what it’s called, you won’t save it) then you’ll be able to use the “Resync” button to reload your changes from for xml files (from memory, it’s in the “Aircraft Selector” window, but I’m not sure). I will post more precise instructions if needed when I’m home at my MSFS PC.

I think you’re talking about the Quick Reload function on the Behaviours window.

But can I do that for an already completed aircraft? I’d need to convert the 307 into a project and can that even be done with all the encrypted files I haven’t recreated? I can’t even properly debug the engines, because dev mode tells me they are encrypted.