investigating Cannot calibrate Gyro with Mag Compass

[Nealius]

I'm having issues with the magnetic compass and gyro heading indicator disagreeing by 15~20 degrees on ramp start, and even after calibrating the gyro heading indicator to match the magnetic compass, after takeoff they disagree by 15~20 degrees again, as if my calibration never happened.

 

The specific conditions are:

- cold ramp start

- Channel map (High Halden)

- mission date April 1, 1944

 

At first I assumed it was a magnetic declination issue, but the magnetic declination of Kent is a smidge over a half a degree, not 15~20 degrees. The second, and more worrying issue, is that the gyro calibration doesn't appear to hold at all. 

 

 

[Habu_69]

Realize this is a necro topic, but I just started flying the Mustang on Normandy 2.0, 2.4.8.39731 OB and noticed this issue of compass and directional gyro readings differing by 15 -20 degrees. Is this a long-standing bug or am I doing it wrong? Do I need to calibrate gyro before takeoff?

[kablamoman]

2 hours ago, Habu_69 said:

Realize this is a necro topic, but I just started flying the Mustang on Normandy 2.0, 2.4.8.39731 OB and noticed this issue of compass and directional gyro readings differing by 15 -20 degrees. Is this a long-standing bug or am I doing it wrong? Do I need to calibrate gyro before takeoff?

Your gyro is dumb and needs to be regularly set to match your compass. It doesn't sense direction, it can only sense your aircraft's rotational acceleration, and it uses that to show you a change in heading. You have to set it to match something (your compass), and it'll constantly drift out of sync from where you set it. Even if it could hold a heading perfectly, the earth rotates around it and will cause it to read inaccurately eventually. To give you an idea, DGs can be expected to drift 3 degrees every 15 minutes, even working perfectly.

To make matters worse, the more you maneuver, the faster it will drift. So, if you're trying to navigate and turn to accurate headings, it's probably best to fly straight and level first to get an accurate reading from your compass, then match your DG (don't forget to uncage it once you've set it) before you use it to turn and roll out accurately at your desired heading using the gyro. 

 

 

Edited May 11, 2023 by kablamoman

[=475FG= Dawger]

Although I seriously doubt its modeled in DCS, the apparent drift of a gyro compass at 48 degrees North Latitude is 11.5 degrees per hour or a degree every 5.2 minutes.

Also, at low power settings, the vacuum source for the instrument may not be providing enough suction to keep the gyro spinning at the proper speed. Also, who knows if this is modeled.

I wouldn't bother setting it until after takeoff and be sure you are in straight and level flight when doing so. Even then it is still going to drift and we really have no clue how this is modeled.

However, pretend it is a worn out gyro subjected to much abuse and its behavior is entirely realistic. I seriously doubt many of them actually worked in 1944 after a few sorties.

[Habu_69]

Ok, but.... see this screenshot shortly after takeoff from Brazenville runway 6. Compass reads 090 and DG reads 060. 30 degrees difference??

[kablamoman]

6 hours ago, Habu_69 said:

Ok, but.... see this screenshot shortly after takeoff from Brazenville runway 6. Compass reads 090 and DG reads 060. 30 degrees difference??

The two instruments are not connected. You are the one that has to match the DG to what you're reading on the repeater compass. The DG could be working just fine but be out 180 degrees if you haven't set it properly.

 

If you notice it out 30 degrees, that's when you fly straight and level (so the compass reads accurately), press the knob below the DG, twist it until the numbers match, and then pull the knob out again.

[Habu_69]

Oh yes, I can quite easily adjust the DG to match the compass indication. But then both instruments will read 090, or 30 degrees from my true heading of 060. I have found no procedure to correct the compass indication to 060. Am I missing something? Surely I am not the only P-51 pilot to experience this issue. How can I ever find my way home from a mission? Perhaps I need to perform a DCS repair.

[kablamoman]

24 minutes ago, Habu_69 said:

Oh yes, I can quite easily adjust the DG to match the compass indication. But then both instruments will read 090, or 30 degrees from my true heading of 060. I have found no procedure to correct the compass indication to 060. Am I missing something? Surely I am not the only P-51 pilot to experience this issue. How can I ever find my way home from a mission? Perhaps I need to perform a DCS repair.

If your compass is showing the wrong magnetic heading that is its own issue and has nothing to do with the DG.

I'd double check your mag heading by using the info bar in an external view to check -- if I am remembering it should tell you what your actual heading is (magnetic and true), make sure you're flying straight and level and then cross check that with what's being shown on the compass. If it's still out to lunch it's a bug with the compass, I would say.

[Habu_69]

OK. Full repair of DCS resulted in no change to the compass behavior. Attaching a track and map screenshot of runway lineup clearly showing my 060 heading. ED needs to look into this or someone advise me what I am doing wrong.

P-51 Compass.trk

[kablamoman]

@Habu_69

I just had an opportunity to go check on this. I can confirm my compass was correctly reading 065 degrees magnetic, flying straight and level, in line with runway 6 at Bazenville on the updated Normandy map. This was exactly correct for the runway, based on the course measuring tool in the map view.

However -- and I think this is what's tripping you up -- when on the ground the compass reads inaccurately and shows 090 as you describe. Ostensibly, this is because the fluxgate in the wing (described as an electronic remote compass unit in the manual) the instrument uses is still susceptible to inaccuracies due to the vertical component of the magnetic field, similar to magnetic dip in an old-fashioned whiskey compass. This is because the fluxgate is no longer flat relative to the horizon as the aircraft's pitch attitude is nose up when on the ground, and so the vertical component of the field lines introduces error.

Long story short: You have to be flying, straight & level -- not in a climb, descent, or a bank -- to get an accurate reading on the compass in order to set your DG.

Edited May 14, 2023 by kablamoman

[kablamoman]

An addendum: Some of these types of systems used gimbals or fluid to help stabilize the sensing units relative to the horizon. It is unclear to me if the unit in the p51 had this kind of functionality (I think some in the German birds used a gimbal stabilized sensing unit, along with some in American bombers), or if it was less sophisticated and would have been susceptible to this type of error as depicted. The behavior in the game at attitudes other than straight & level may not be entirely correct, but I'm guessing that's how they've chosen to model it, and so flying straight & level is what you have to do to get an accurate reading.

Edited May 15, 2023 by kablamoman

[kablamoman]

@NineLine After a bit more searching it looks like the p51 remote compass sensing unit was indeed suspended in fluid to correct its attitude relative to the horizon. The inaccurate readings due to the vertical component of the magnetic field lines when pitched up, down or banked relative to the horizon (easy to demonstrate by flying a heading, and then pitching up or down and holding that attitude, or simply being wheels down on the ground) do not seem like correct behavior for this instrument and may be worth looking into. 

Some references attached:
 

 

 

 

Edited May 14, 2023 by kablamoman

[=475FG= Dawger]

@kablamoman

The parts catalog you posted lists it as a remote indicating magnetic compass, which isn’t a flux gate compass. If it truly is a magnetic compass mounted remotely it would still suffer from all the errors associated with a magnetic compass. If it is a misnamed flux gate, then it is likely modeled incorrectly.

From reading the parts list, I’d say it is a magnetic compass and ED has the modeling essentially correct.

[kablamoman]

@=475FG= Dawger

That's what I had been assuming as well, but after poking around a bit and reading the manual description:

Quote

"This type of compass doesn't float around and fluctuate when the plane maneuvers, and it gives you all the advantages of the directional gyro without precession inevitable in the directional gyro. Both are provided, however, the directional gyro being available in an emergency should the electrical system fail."

I started to wonder, and when looking through some documents, this popped out to my eye in the parts list for the sensing unit:

(unfortunately I did not have the entire manual)

However, your comments made me even more curious, so I actually paid to download the full copy of it and here are some of the missing pages:

So I think it's at least worth a closer look.

Edited May 14, 2023 by kablamoman

[Habu_69]

Very impressive research! My track shows the 30 degree error continues to exist during initial climb-out after takeoff, but of course I am pitched up during that maneuver. I will check out behavior when straight and level.

[kablamoman]

It's still unclear to me if the flux gate in the assembly is used to actually sense the earth's magnetic field (suspended from the float as a pendulum to keep level), or if it's being used simply to generate electric signals from the float, which would be doing the sensing. I am leaning towards the latter.  In any event, some additional information that may be helpful comes from a partial manual I found that describes the testing procedure, specifically the tolerances for tilt error on the AN 5730-3 remote sensing unit:

 

The section mentions that a tilt about any axis of 10 degrees should not result in a reading change of more than 6 degrees.

The ~25 degree error when sitting on the ground (about 12 degrees, nose up), as in OP's original example, really does seem to be excessive.

Edited May 14, 2023 by kablamoman

[kablamoman]

@=475FG= Dawger

On 5/14/2023 at 4:44 AM, =475FG= Dawger said:

If it truly is a magnetic compass mounted remotely it would still suffer from all the errors associated with a magnetic compass.

So just assuming maybe they've modeled it as a traditional wet compass... Playing around with it a bit more (latest MT build) it becomes really pretty confusing. It looks like the errors in question don't behave like traditional magnetic compass acceleration errors at all. That is to say, there's an error element based solely on tilt angle even at zero acceleration. You can be stationary on the ground, or unaccelerated while holding a pitch attitude (nose high or low) and see that the error is directly correlated to the deck angle. A wet compass should only really be inaccurate for periods of acceleration, such as in a turn or during speed change -- these situations will both displace the fluid dampened pendulum or float device from level relative the earth -- but once the float stabilizes level to the horizon (eg. on the ground; in an established, unaccelerated climb or descent; level, but at low speed with a nose-up attitude; etc.) the errors should largely subside. Unless of course your attitudes are sufficient enough that the physical casing starts to impede free rotation -- but as mentioned in the above document, the float in this case, should be freely rotating up to +-20 degrees of attitude change.

It's as if they've modeled it as a free-wheeling sensor that rotates freely about the normal/vertical axis (and will happily seek out magnetic north, if it's steeply inclined along this plane of rotation), but they haven't modeled the "float" part, that should largely correct for inclination at modest deck angles.

Regardless of sensing aspect of the instrument, I don't see how this behavior can be accurate.

Edited December 22, 2023 by kablamoman

[=475FG= Dawger]

Reading all that, I think its a flux gate compass and ED has the modeling wrong.

Nice job digging all that up.

 

 

[303_Kermit]

Here's something about gyro. I hope it helps a bit. Sperry gyroscopes.
https://www.maritime.org/doc/gyromk14/index.php

[NineLine]

Hey guys, sorry for not responding sooner. If this is still an issue can you give me a summary of the problems decided upon in your discussion above and I will dig into it.

Thanks!

[kablamoman]

10 hours ago, NineLine said:

Hey guys, sorry for not responding sooner. If this is still an issue can you give me a summary of the problems decided upon in your discussion above and I will dig into it.

Thanks!

Yes, still an issue.

The magnetic compass is susceptible to excessive “tilt” error any time the deck angle of the aircraft is not level with the horizon.

In the Bazenville example near the top of the thread, when aligned with runway 6, instead of reading 65° magnetic, heading on the compass can be out as much as 25 degrees (90° magnetic) with a slight pitch up attitude.

The compass incorporates a pivot and float device that should rotate freely when tilted and should not be so drastically affected by inclination/dip errors (see the remote sensing unit test tolerances for tilt error I posted here).

Edited November 8, 2023 by kablamoman

[NineLine]

Thanks, I am gonna try and take a look this weekend. 

[kablamoman]

On 11/10/2023 at 10:43 AM, NineLine said:

Thanks, I am gonna try and take a look this weekend. 

Thanks @NineLine.

Any word on this? I just did some more testing and it looks like it's an issue related only to the modules using a remote compass installation.

The P-47 and Spitfire correctly read 65 degrees magnetic sitting on the ground in Bazenville aligned with a magnetic heading of 65.

The P-51 and all Axis birds read ~90 degrees with their deck angle titled up on the ground, and the error seems to modulate in relation to the deck angle. Flying with a neutral deck angle along the same heading results in correct readings.

Screenshots attached:

P-47 with correct reading:

P-51 and all axis birds showing 90 degrees:

 

 

As mentioned, the P-51 float device is suspended in fluid with a pivot that allows it to maintain level with the gravity vector (horizon) until the casing itself starts to impede free rotation, and as far as I know the Luftwaffe "Mutterkompass" installation in the back of the Axis birds was mounted with gimbals allowing it to stay level with the horizon through a range of deck angle changes, and so deck angle of these aircraft should not lead to their remote sensing compasses being affected by magnetic inclination (the vertical component of the magnetic field) the way they appear to be at moderate deck angle changes away from perfectly level.

 

Edited December 20, 2023 by kablamoman

[kablamoman]

A little update with some more clarification on the "flux gate" shown in the diagrams, in case anybody is interested. It is not a flux valve compass like in a more modern installation, but a simple toroidal coil used to sense the position of the compass float. "Magnesyn" was a patented method for converting mechanical movement to electrical signals and back again for remote indicating instruments.

The remote compass in the P51 is indeed a traditional wet compass, it just happens to be located in the wing (to be free from any magnetic interference in the cockpit).

Edited December 22, 2023 by kablamoman

[JBDCS]

I don't know if this is of interest, but here's  a document I found when doing an internet search for info on the repeater compasses.  It's a report from 1945 "Review of Remote Indicating Systems for Aircraft", covering various R.I. devices including compasses.  I haven't read it in any detail, but it looks like there's some information on how flux gates were used at that time in it.

https://reports.aerade.cranfield.ac.uk/bitstream/handle/1826.2/3199/arc-rm-2199.pdf?sequence=1&isAllowed=y

[Edit: On having a look at it and searching for "flux gate", it's only mentioned with respect to one particular transmission method.  Whether that's relevant to the repeater here I don't know.  In fact it looks like it is, since it is the Magnesyn transmission method mentioned by kablamoman.]

(Found this document, which is in the UK Cranfield University Aerade archive, via a link from a radio forum thread which the search turned up: https://www.vintage-radio.net/forum/showthread.php?p=1488845  Looks like there may be other interesting stuff in the Cranfield archive, but I've not had chance to look at it yet: https://reports.aerade.cranfield.ac.uk/)

Edited December 27, 2023 by JBDCS
Fixed garbled URL and added comment after further reading