investigating Artificial Horizon Intermittent

[Land0]

I've noticed that the artificial horizon is often not aligned with the horizon. Usually because it does not erect properly after engine start and remains toppled.

[PETER SHIFTY]

Right i have the same problem 

[Bozon]

I was wondering the same thing. We also can’t cage the artificial horizon to reset it like in the other warbirds.

[scoobie]

I can't check now, but... AFAIK that's how this British horizon worked (Model Something Mark A Few, can't remember off the top of my head). The same or similar instrument is in the Spitfire. This thing doesn't have cage/uncage knob for the pilot, but settles itself over time. Pretty long time, not seconds. I haven't checked in the Mossie, but maybe an experiment could help: first have your horizon "skewed" for whatever reason, disturbed, then enter straight & level flight and keep her like that for a few minutes. Does the horizon "straightens & levels" itself? If so, apparently it was meant to work like that.

Yesterday I was cruising around neighbourhood just for the pleasure of cruising around and noticed that the horizon was perfectly straight and level. So maybe it just works?

 

American planes had different horizons... and they were pretty bad, horizons that is, not planes - at least as they are depicted in DCS. On the other hand, these planes had sensitive slip & turn indicators, so at least for cruise you were supposed (I think) to use primarily them (plus altimeter and VVI) - they're excellent for that job.

AFAIK, artificial horizons were kind of new tech at that time and were far from perfect, though I stand to be corrected on this.

 

[JetJake76]

5 hours ago, scoobie said:

This thing doesn't have cage/uncage knob for the pilot, but settles itself over time.

 

 

Thanks for the tip, I will check it today.

[scoobie]

Hmmm... just did this experiment myself. Knocked the gyro and went flying straigh & level for over 10 minutes. It didn't erect, I'd say it got even more messed up than it was on the take-off (screenshot attached... sorry for not taking initial screenshot, I forgot).

 

So umh... anybody knows how to erect a knocked gyro in the AI (AI for attitude indicator) if you don't have the cage knob?
Maybe there's no vaccum, I need to flick some switch, ask nicely or something?

 

[scoobie]

Let others speak for me.
(AI = attitude indicator)

 

Firstly an excellent computer animation of pendulous vanes:

 

 

Secondly how it works in a real AI (fast forward to about 1:43):

 

Thirdly an old picture. It may be from WW2 or later, I don't know, but it definitely shows an EARLY design.
Source: https://commons.wikimedia.org/wiki/File:ArtificialHorizonVacuumGyroErectionMechanism.jpg

Pay attention to how the vanes look like:

 

Lastly the British Mk. IB (Spitfire) AI.
Source: https://caggius.wordpress.com/artificial-horizon-mk-1b-6a-1519-teardown-spitfire/

The below picture comes from this site, but I added annotations in red:

So, yeah, I understand there might have been little motivation to do self-erection simulation in DCS if (so I heard) the maps in DCS are flat (no Earth's curvature). And to an extent you may say it was fine for as long as all warbirds had either a cage knob (USA) or a cage ring (Germany). In these planes, if your AI goes bad, you're annoyed and as soon as you can fly straight and level, you cage/uncage and you're good to go again.
But how about the Spit and now the Mossie? Once you upset your AI, it's scr*wed until you restart the mission, so it seems.

 

 

Edited September 23, 2021 by scoobie

[Jakey-Poo]

I made a topic about this for the P-47, and I have a suspicion that there might be an issue with the the model(s) themselves, specifically inadequate suction in the vacuum system.  In the P-47 during cold start you can clearly see that the suction never reaches adequate range for nominal operation per the gauge.  This may be the same issue for the mossie.  I'll attach that post below.

 

 

[SMH]

I've always wondered this about the Spit and now Mossie too. And yes, the AH seems even more unstable in the Mossie when of course they should be the same. (Not sure which one is right. Hard to imagine wasting space on the panel for an instrument this useless though.)

[NineLine]

Track? We need tracks guys.

[scoobie]

Track included.

HEALTH HAZARD: EXTREME BOREDOM!
Events in the track file (AH = Artificial Horizon):
1. Cold & dark, but AH is erect (side note: I wonder how ... but we're not talking about this here).
2. Start her up, AH is still erect (OK).
3. Shut her down, AH slowly "collapses" (OK).
4. Start her up again, set boost +7 @ about 2,750 RPM and see what happens next.
5. AH is slowly collapsing even further! (not OK, I assume)
6. Perform a considerably sloppy take-off to present that the gyro in the AH must be spinning, because the AH is basically working - it's just skewed with the error already accumulated, but spinning. Unfortunately in the track file I nose down and crash some 10 seconds after the take-off, which wasn't really the case, but it doesn't matter - before it happens you can see the AH is moving. Anyway, I hereby testify that it doesn't erect in flight either.

 

ah_no_self_erect_mossie.trk

Edited October 5, 2021 by scoobie

[Kiwi_Viper]

same here, cant find any tracks tho its not creating them did get screenshot

[YoYo]

The same, from cold and dark start. If you start from runway or in the air will be ok, looks like a bug.

[NineLine]

Thanks guys, taking a look today.

[YoYo]

20 minutes ago, NineLine said:

Thanks guys, taking a look today.

Thx, here is example (from 30 min.): 

Reflected's video, the same.

However what I noticed after 10-15 minutes of my flying it somehow aligns himself (so maybe feature)? To confirmation.

[scoobie]

@YoYo, are you sure?
That would mean the artificial horizon's gyro self-erection actually works and we were all wrong.
You know... there's a secret contest for a bottle of good whiskey: whoever shows a track or video where the horizon actually self-erects, gets the bottle

 

[Robi-wan]

On 11/1/2021 at 1:00 PM, NineLine said:

Thanks guys, taking a look today.

Any progress to report regarding the attitude indicator? 

On 11/8/2021 at 5:35 AM, scoobie said:

That would mean the artificial horizon's gyro self-erection actually works and we were all wrong.

I'm still experiencing a completely u/s attitude indicator when starting with a cold/dark airplane. Even with 15+ minute legs of wings level, unaccelerated flight the attitude gyro does not erect.

[DragonSoulkin]

Problem still persistent in the P-47 and P-51. The AH will get all wonky just sitting on the ground during warmup after uncaging. I had the track file but it said it's too large to upload (5mb size limit) it's like 28mb.

For the most part this is just an inconvenience, until I'm flying through cloud cover or in a storm and I can't tell up from down besides watching my side slip indicator and altimeter.

Edited November 24, 2022 by DragonSoulkin

[gcze]

Same for me, artificial horizon almost useless. This it's huge problem where flying at night. 

It looks like a problem with propeller torque effect.

Track in attachment

test.trk

[NineLine]

Thanks all, will take another look. 

[-0303-]

Artificial horizon doesn't work correctly on a single airplane AFAIK. I and others have tested a dozen or more planes. There's an exhaustive thread on this since April 20, 2021.

It boils down to the pendulous vanes function is never implemented or bugged. Seems this bug was and is copied to every new plane (like the Mosquito). Artificial Horizon is part of the "sixpack",  the six basic flight instruments.

 

 

Was gearing up for a yearly anniversary bump to that thread. Got an acknowledgement of this for the Spitfire and have waited for this to be addressed (for Spitfire and every other plane).

Edited January 6, 2023 by -0303-

[ldnz]

Couple of things I found in V for Victory so far:

Because its cold, I find the engines take a lot more priming to start. If I don't let them prime enough, and the first engine coughs and dies, then the AH is stuck oriented at whatever position it briefly got to. As you change course, that apparent angle will change as it seems fixed in a vertical position.

If I prime the engines heaps and get a clean start, then the AH is ok. So it seems like as soon as first loss of vacuum occurs, its irretrievably broken?

[ldnz]

Belay last - I can't find any reliable technique for the AH. Hopeless. The DI is also swinging wildly with engine RPM < 1500 which makes no sense to me - if its a gyro instrument, why would it swing back and forth with vacuum fluctuations? Surely if vacuum is low it would be more prone to drift, but thats a slow process.

[-0303-]

It makes no sense vacuum intermittence having anything to do with it. Big picture, there's a simple test that every plane is failing:

• If the AH becomes misaligned for any reason, just flying straight & level should realign it within a reasonable time.

Video shows a real Spitfire takes 9 minutes to self correct from a 90 degree misalignment. Watch 18:05 - 27:00. 360 video, just tilt down a little to see the AH.

Every plane with a caging button (ex free TF-51D) is easy to deliberately misalign for the purpose of this test. Bank, cage, roll back level and uncage. Now it's misaligned, fly straight and level and watch it (not) self correcting. Neither Spitfire nor Mosquito have caging buttons.

There's another test that also fails:

• Doing rolls and loops should scramble the AH because of the mechanical stops inside the AH.

Again, the real Spitfire video shows this. Every roll the AH goes coco (why I noted timestamps for every roll). DCS doesn't do this.

 

 

[Hog_driver]

Apart from a properly working AH, could we get a "quick & dirty" solution such as, let's say, "press X to reset the AH"?