How much INS deviation can we expect on a 500nm enroute or 45 mins of flight with GPS on?

[Tholozor]

AFAIK, the 300ft metric is applicable regarding horizontal position error. ACAL for vertical error provides either GPS-only, DTS-only, or BOTH options for vertical position (BOTH being the default, and the MMC picking whichever of the two is more accurate based on specific parameters). ACAL-provided vertical position error shouldn't ever be more than 100 feet from what I'm reading (unless a system has failed).

[Mapi]

no fun

[darkman222]

I did the same test again. This time for INS drift in altitude. Flew 600 nm in one direction and back. Total travel distance 1200 nm.

I had a Bedford truck at the runway end with a waypoint co-located. When I came back and landed the INS drift in altitude was (considering the height of a Bedford Truck) in about 30 ft higher than the waypoint was when I took off. So the INS drift in altitude is also well in the allowable limits and negligible.

What I still cant figure out is while the INS dift in all axis seems to be in the limits, how is it possible to create the impression that the waypoint is off more than in the limits of 300 ft on the ground plane when approaching in a shallow angle. As seen two posts earlier.

Edited June 27 by darkman222

[LastRifleRound]

The bigger problem here is why isn't FIX and cursor slew working correctly? No matter how large or small the drift is supposed to be, it cannot currently be corrected. This is a pretty big problem and certainly not intended. MOVING THE CURSOR should update the whole flight plan in an F16. That's why Cursor Zero is so important to understand in the Viper. FIX just resets the cursor zero point to the cursor's current position (be it radar, HUD, TGP, etc), but slewing the cursor is what moves the flight plan. The current Viper does not do this despite the fact that this behavior is well-documented in the dash manual.

Edited July 1 by LastRifleRound

[darkman222]

On 7/2/2025 at 1:55 AM, LastRifleRound said:

FIX and cursor slew

What do you mean? I tried it and it works. If you have lets say, 4 waypoints close to each other, you cursor slew one, all the others are moved accordingly. CZ and they are all moved back to the initial position. No issue here. 

On 7/2/2025 at 1:55 AM, LastRifleRound said:

but slewing the cursor is what moves the flight plan

If you use FIX, it does  the same. But you can not CZ to the initial position any more obviously. 

So can you be more precise what you are referring to?

Edited July 6 by darkman222

[LastRifleRound]

Was it fixed recently? Wasn't mentioned in any notes. It was not doing that before.

[darkman222]

Was fixed and mentioned in : en/news/changelog/release/2.9.17.11733/

->Fixed: INS drift is not GPS corrected and INS drift adds up to more than 1000 ft over time.

Which is the exact title of the bug report that I submitted about it in the bugs section.

 

Its now in the allowable limits. Under some circumstances the drift seems higher though. But the reasons will remain a mystery, I guess... Thats why I did the additional test of the drifting in altitude. Which is also in the allowable limits. 

[LastRifleRound]

That says it drifts less, not that fixes are working

[darkman222]

From what I understand is that our (pre EGI) DCS F16 will always have an induced INS drift, that will be corrected via GPS within the limits, which is happening now after the patch has been released. 

So what was fixed is not that it does not drift anymore, because this is not realistic. What has been fixed is that the drift stays within the allowable realistic limits. Thats what I tested here, and which is the case. As long as we agree on a altitude drift of 30 ft ( 100ft is allowable) and the lat/ long drift is about 300 ft, the fix is on point. Although 300 ft should be the maximum drift where GPS should start correcting it. 

If you read through the entire thread discussion which is not too long, 300 ft is the maximum drift, when GPS should be activated to start to correct it to a nominal drift of 131 ft (40 meters) . Which is also the amount of drift that is talked about in the Viper Mini Updates white paper, dealing about Kalman filter. If we want to go on discussing something here then it would be the only thing that still stands out: If it is 300 ft or 131 ft we should be experiencing with GPS corrections active.

Edited July 14 by darkman222

[itn]

On 7/14/2025 at 2:38 AM, LastRifleRound said:

That says it drifts less, not that fixes are working

FIX has been working for years. 

However, if GPS is enabled, it over time would ”override” your manual FIX. Some people have observed this, but haven’t always realized what exactly happens, and then have claimed ”FIX does not work” here in forums. In reality it’s often just that GPS overrode the FIX they did.
 

In essence, at least some versions back doing FIXes in GPS environment was pretty useless. In non-GPS environment they’ve worked just fine. 

[darkman222]

But what would you do then, if FIX is overridden by GPS. When the accuracy is really 300 ft only instead of 131 ft, which has not been answered yet?

300 ft of INS drift that GPS does not fully correct but that can not be corrected by FIX taking either?

Edited July 16 by darkman222

[Muchocracker]

On 7/14/2025 at 8:47 AM, darkman222 said:

From what I understand is that our (pre EGI) DCS F16 will always have an induced INS drift, that will be corrected via GPS within the limits, which is happening now after the patch has been released. 

So what was fixed is not that it does not drift anymore, because this is not realistic. What has been fixed is that the drift stays within the allowable realistic limits. Thats what I tested here, and which is the case. As long as we agree on a altitude drift of 30 ft ( 100ft is allowable) and the lat/ long drift is about 300 ft, the fix is on point. Although 300 ft should be the maximum drift where GPS should start correcting it. 

If you read through the entire thread discussion which is not too long, 300 ft is the maximum drift, when GPS should be activated to start to correct it to a nominal drift of 131 ft (40 meters) . Which is also the amount of drift that is talked about in the Viper Mini Updates white paper, dealing about Kalman filter. If we want to go on discussing something here then it would be the only thing that still stands out: If it is 300 ft or 131 ft we should be experiencing with GPS corrections active.

The GPS and INS are 2 disconnected systems operating concurrently. The Mission Computer will initially just output a position estimate based on the INS alone, but it is observing the position errors between it and the GPS. When it recognizes that the INS exceeds 300 feet of error it will then introduce the GPS data into the kalman filter and correct the blended positiong estimate in the MMC. There is no actual automatic fixing going on in the INS like you would see with EGI. For the kalman filter, the GPS is just there to be a reference (a noisy one) that it uses to corral the blended position estimate to under 300 feet. 

When running a FIX it should then correct the INS back to an accurate error state. The kalman filter will at that point stop blending the position estimates and go back to trusting the INS alone again, until the error accumulates back up to above 300 feet and the cycle repeats. I have not flown the F-16 in quite a long time, but that is my recollection of how the FIX function should operate.