Doubt with P-38 J-25 directional gyro

[LF_Mark_Krieger]

I have a doubt with P-38 directional gyro. I noticed it doesn't work on the ground, while magnetic compass and remote indicating compass do. Probably this is normal. But I noticed too that while flying, when the heading is 0º or 180º it is precise, but not around 090º or 270º. Then it acumulates error. Is this normal? Im doing slow turns, so the gyro doesn't suffer a sudden error. The pics are taken in stable flight.

  

Edited June 11 by LF_Mark_Krieger

[LF_Mark_Krieger]

I usually don't like to ask to the AI, but this look like a reasonable response:

Yes, the directional gyro (DG) of the P-38J-25 Lightning, like all gyroscopic heading indicators of the era, accumulates error over time, regardless of the heading — east, west, north, or south. However, errors accumulate more rapidly when flying east or west due to a phenomenon called transport error (a subset of precession errors).

Why this happens:

Gyroscopic instruments like the directional gyro are based on the principle of rigidity in space — they maintain a fixed orientation in inertial space. However, the Earth is rotating, and when an aircraft moves across the surface, especially east or west (along lines of latitude), it causes a misalignment between the gyro's reference and true geographic direction.

Specifics for east/west flight:

• When flying east, the Earth’s rotation makes it seem like you're turning slightly to the left (from the gyro’s perspective), causing the DG to slowly drift.

• When flying west, the opposite happens — the gyro seems to drift to the right.

• This effect is more pronounced the farther you are from the equator and the faster you're traveling east or west.

For the P-38J-25:

The aircraft didn't have a slaved gyro or automatic correction. Pilots had to regularly realign the DG using the magnetic compass, typically every 15 minutes, to counteract these accumulated errors.

Summary:

Yes, the DG in the P-38J-25 accumulates error when heading east or west, more so than when flying north or south, due to the Earth's rotation and the resulting transport precession error.

[MDzmitry]

Has the problem remained after flying level for a while? Similar stuff can be observed when flying the Spit, there one needs around 10-20 seconds for the gyro to get its bearings following a sharp turn.

[MaxPower]

I don't think this AI summary is explaining why the error is only accumulating for certain directions and not for others.

[czech693]

IRL, the directional gyro compass is used for navigation because it doesn't bounce around or lag the turn like the magnetic compass.  I can't tell if you are turning in that top photo.  The directional gyro has to be adjusted to the magnetic compass heading periodically as it will drift (IRL).  Since there's no way to adjust the directional gyro, I assume the sim will eventually adjust it, otherwise it's useless after it drifts off.  As to the sim having it exactly on in the north/south alignment, but off in the east/west, I can't explain as it is not magnetic and does not point to magnetic north.

[MaxPower]

I think you're trying to explain a game bug by referencing how real aircraft systems work.  I think gyro error accumulation, or so-called drift, is an offset, and should not affect one direction more than others.

When you're coding rotation calculations, all kinds of things can go wrong.  In fact, it's so hard, a common method is to use quaternions which invokes a forth imaginary dimension to try to explain rotations in three dimensions.  Then you have problems with nested transforms and aligned axes causing 'gimbal lock'- it just seems to me like there's a lot that can go wrong in the execution of the rotation animation that has nothing to do with how real gyros behave.

Edited June 11 by MaxPower

[LF_Mark_Krieger]

22 hours ago, MDzmitry said:

Has the problem remained after flying level for a while? Similar stuff can be observed when flying the Spit, there one needs around 10-20 seconds for the gyro to get its bearings following a sharp turn.

Yes, the error persists. The game simulates the pilot correcting the gyro some seconds after doing a sharp turn, because the gyro cannot stand sharp turns. This happens in all gyros.

12 hours ago, czech693 said:

IRL, the directional gyro compass is used for navigation because it doesn't bounce around or lag the turn like the magnetic compass.  I can't tell if you are turning in that top photo.  The directional gyro has to be adjusted to the magnetic compass heading periodically as it will drift (IRL).  Since there's no way to adjust the directional gyro, I assume the sim will eventually adjust it, otherwise it's useless after it drifts off.  As to the sim having it exactly on in the north/south alignment, but off in the east/west, I can't explain as it is not magnetic and does not point to magnetic north.

In that photo the flight was stable. The problem is the game only adjusts the Gyro some seconds after doing a sharp turn. the major error flying west or east is due to the rotation of the earth, it seems.

What I think it's not well simulated is that the game uses an automatic error since the first second, when in reality this error was slowly getting accumulated through time (faster if you were flying east or west, but in a progressive way). Perhaps it would be ask too much to the developers to find a system that simulates this error in a more similar way to reality? I don't use the hud compass, but I imagine most players don't deactivate it and don't even look at the gyro. 🤔

[MaxPower]

Ooooh.  I think I misunderstood the original post.  I thought you were doing turns, and there was no error when you turned north or south, but there was error flying east or west.  So in the same flight, pointing north or south would show no error then pointing east or west would show some error.

[LF_Mark_Krieger]

4 hours ago, MaxPower said:

Ooooh.  I think I misunderstood the original post.  I thought you were doing turns, and there was no error when you turned north or south, but there was error flying east or west.  So in the same flight, pointing north or south would show no error then pointing east or west would show some error.

Exactly. I just said my turns were slow, because the gyro gets decalibrated in sharp closed turns. I already knew that, but I didn't know that it accumulates error anyway during flight. What I am noticing now it's that it is not well simulated or it is a bug. I think that in other planes with gyro the error isn't so clear. This evening I'll have a look and compare with other planes with gyro.

[czech693]

IRL a directional gyro turns smoothly during turns.  That's why it's used for navigation instead of the magnetic compass which swings and lags during turns.  With the directional gyro you can stop the turn exactly on the heading you want.  A turn in any direction should not cause it to lose it'settings.  It will gradually drift off over time.  In training you are taught to check it against the magnetic compass and re-adjust it back to the magnetic heading every 15 minutes.

[LF_Mark_Krieger]

On 6/11/2025 at 3:28 PM, czech693 said:

IRL a directional gyro turns smoothly during turns.  That's why it's used for navigation instead of the magnetic compass which swings and lags during turns.  With the directional gyro you can stop the turn exactly on the heading you want.  A turn in any direction should not cause it to lose it'settings.  It will gradually drift off over time.  In training you are taught to check it against the magnetic compass and re-adjust it back to the magnetic heading every 15 minutes.

So it is something that isn't simulated correctly. What a pity. 😔