MSFS Attitude Indicator Issues

[Zoren]

I have converted my FSX/P3D project into MSFS and have most of the animations working. My one issue is with the attitude ball animation. I updated the xml I used within P3D to the newer MSFS one using the "Here's how to write you own ModelBehavior code." found within this site. Most of my other gauges are working pretty good except the attitude indicator. At first it was animated like the P3D one, 0-360 for both pitch and bank (values - 180/90/0/-90/-180 // -180/-90/0/90/180 set on frames 0-90-180-270-360). This works fine in P3D using the below code from modeldef.xml but when used in MSFS the bank goes all over the place with it sticking/stopping/going. Sometimes it will roll nicely until it reaches back to level, stops for a bit and then continues if you continue your roll.

<UseTemplate Name="ASOBO_GT_Anim_Code">
   <ANIM_NAME>attitude_indicator_bank</ANIM_NAME>
   <ANIM_CODE>(A:ATTITUDE CAGE,bool) 1 &lt; if{ (A:ATTITUDE INDICATOR BANK DEGREES, degree) 180 + } els{ 45 }</ANIM_CODE>
   <ANIM_LAG>0</ANIM_LAG>
   <ANIM_LENGTH>360</ANIM_LENGTH>
</UseTemplate>
<UseTemplate Name="ASOBO_GT_Anim_Code">
   <ANIM_NAME>attitude_indicator_pitch</ANIM_NAME>
   <ANIM_CODE>(A:ATTITUDE CAGE,bool) 1 &lt; if{ (A:ATTITUDE INDICATOR PITCH DEGREES, degree) } els{ 45 }</ANIM_CODE>
   <ANIM_LAG>0</ANIM_LAG>
   <ANIM_LENGTH>360</ANIM_LENGTH>
</UseTemplate>

Does anyone have an example of their attitude indicator they would like to share or help with this? Maybe my animation number in Blender are wrong.

 

[kalong]

delete your <ANIM_LAG> for smooth loop roll

 

[B21-soaring]

I agree there seems to be something very strange with the ATTITUDE INDICATOR vars. On my aircraft the BANK value seemed to work at first in a steady turn, but then the value wobbles in the direction of the turn and become meaningless.

The only guess I have is some error condition (insufficient power?) is kicking in.

My gauge is html/js and I changed from Simvar.GetSimVarValue("ATTITUDE INDICATOR BANK DEGREES", "degrees") to

let xyz = Simvar.GetGameVarValue("AIRCRAFT ORIENTATION", "XYZ")

( *var name from memory - typing this on a phone)

which then gives the required values as:

xyz.pitch, and
xyz.bank (radians)

With just that straight swap of the vars, my horizon gauge immediately worked perfectly, so there's something fishy with the other simvars.

 

[B21-soaring]

fyi I doubt you can use the orientation var in the XML, but you could write a 10-line html/js gauge that simply reads that orientation var and writes a couple of L: vars you could use instead in the XML.

Or maybe better, someone here knows the secret of the wobbly ATTITUDE INDICATOR vars.

 

[WarpD]

PLANE PITCH DEGREES and PLANE BANK DEGREES is what I've always used. Despite the name, they return as radians by default. You can, obviously, request it in degrees as you did for the ATTITUDE INDICATOR BANK DEGREES variable.

 

[sal1800]

Can you check the animation keyframe interpolation setting? It should be on linear. I had a similar problem with an altimeter.

 

[B21-soaring]

for my MSFS html/js horizon gauge I can simply display the simvar value and watch it drift. I don't know if the OP has the same underlying issue but my testing confirmed 100% that the attitude vars are unreliable (where the AIRCRAFT ORIENTATION XYZ special var is rock solid). I have not used the attitude vars before so I don't know if they are unreliable by design, simulating some old-school gyro instrument failure?, which is what my drift looked like. E.g. an electricity or vacuum issue?

 

[Roy Holmes]

I would suggest using pitch before bank in your expression. The reason is that pitch goes from plus 90 to minus 90 and bank is continuous.
Roy

 

[WarpD]

Bank is +/-180...

 

[Zoren]

I would suggest using pitch before bank in your expression. The reason is that pitch goes from plus 90 to minus 90 and bank is continuous.
Roy

Not sure how to do that with xml code but should the parent object be pitch and the child bank? Right now I have the bank as the parent and the pitch child.

I'll try that later and see if that works better.

 

[B21-soaring]

I have the bank as the parent and the pitch child.

this seems correct to me, so the rolling motion axis does not move up and down and the pitch moves within the rolling motion. If the animations are parent-child in the model then the order in the <model>.xml is irrelevant (& Roy didn't have that info).

I know I'm repeating this (apologies, this is the last time) but are you sure your problem isn't something like the gyro electricity is actually off ?

Here's a debug suggestion: replace the bank ANIM CODE with this (which counts 0..90 every 10 seconds):

<ANIM_CODE>(E:ABSOLUTE TIME,seconds) 10 % 9 *</ANIM_CODE>

If the indicator behaves as expected (a linear bank rotation 0..90 degrees every 10 seconds, snapping back to 0) then your model & model.xml is OK and there's a problem with the simvar.

 

[Roy Holmes]

If you take an attitude indicator apart or just look at a diagram of what it does, you will see this;
The case is attached to the airplane. Inside the case is a gyro which tries to stay still as the airplane pitches and rolls. The gimbals for the gyro have arms that will move a pitch display up and down relative to it as the airplane pitches. Those arms are held by a rollable gimbal that rotates when the airplane rolls. So the gyro is the parent and it directly drives the pitch display. The display component of the pitch display can also roll according to bank so bank is the child.
Roy

 

[B21-soaring]

Roy I get the real instrument description but I think the extrapolation into sim programming is incorrect. Maybe it can be thought of as an 'inversion' that occurs when the sim code rotates the ball with reference to the airframe, rather than the aircraft rotating around the ball in a real aircraft. By a major coincidence I wrote an Attitude Indicator gauge earlier this week, with bank as the parent, and it's working like a charm.