reported GPS not arresting INU velocity errors

[Crptalk]

After a 2 hour flight I'm stationary on the ground while the helicopter thinks I'm doing 7 knots and descending at almost 1000 fpm.

ah64 inu drift.trk

Edited March 20, 2022 by Crptalk

[Talonfox]

I experienced a similar situation last night. After rebooting this morning, the symbology had returned to normal but after a short hour flight it began to drift again. I attached a screenshot of the condition while I was reloading at the FARP. Continuing to collect data on this anomaly. 

[Crptalk]

Temporary work around I've been doing is to reset and switch INU's during long flights.

Edited March 21, 2022 by Crptalk

[Wychmaster]

Thank you for reporting this. I also experienced the drift. It also affects the stored points. If you select a target with the TADS, store its location, make it the acquisition source and then slave the TADS to it, you will see how the point is moving.

It always happens when I fly the same chopper for longer periods.

Edited March 21, 2022 by Wychmaster

[Scaley]

Has anyone tested to see if the doppler helps? It appears to be off in hot-start aircraft.

[jayst0r]

I have I always enable doppler in my startup routine, it does not help. Sometimes also just reseting the 2 INU is not enough. As they where still shown as good but it had no position.

What then helped was pulling down the throttle, wait a bit, throttle up again then reset INU, then they start aligning again.

Edited March 25, 2022 by jayst0r

[BIGNEWY]

Hi, 

we do have this reported and the team will look into it. 

thanks

[Talonfox]

Here is another sequence of symbology drift with associated timestamps. If needed I can DM the 45 MB track file.

Best of luck to the troubleshooting team.

[Swift.]

The INUs can be observed to experience uncorrected errors in their velocity. 

In this test I sat stationary on the ground with the time accelerated, however similar issues have been observed after a couple of hours of flying as well.

It can be seen that there is a full GPS acquisition with several satellites being indicated. As the mission progresses, the INU position confidence can be seen to fluctuate, and even reach as high as 0.099KM at one point, but is fixed (presumably by the GPS) and returns to nominal. However, the IHADSS velocity vector and vertical speed indicators can be seen to wander around throughout the test, and at the 1 hour mark are noted to be significantly displaced from neutral.

The velocity errors these indicators suggest are different between INU1 and INU2 and can be corrected by a full reset of one of the INUs as demonstrated at 6:45 and 7:40

To summarise, there appear to be velocity errors being introduced into the INUs that are not correctable by any pilot action short of complete INU reset and are not automatically corrected by the aircrafts systems such as GPS or Doppler (whilst in flight).

 

NB. Track not included for the length of time required for this bug to be shown, reproduction is a simple matter of sitting and waiting.

[frostycab]

Regarding the Doppler, this could well be down to the fact that you were sat still and didn't go anywhere. I'm not familiar with exactly how it works on the Apache, but the name suggests the system depends on the principle of Doppler Shift, which is only observable when in motion relative to the signal. Presumably it uses a similar principle to the Rad Alt, and examines the frequency of the returned radio signals to derive the motion on the emitter, passing that onto the INUs to help correct them. When sat still on the ground like you are in the video the emitter would be getting no frequency change and thus has nothing to compare it's position to. Perhaps that might means it should tell the INUs that there's no motion so they stay aligned and centered, but there could be a weight-on-wheels lockout to protect ground crew.

I'd love to hear from someone who knows the system well and can explain. I've been using the radar altimeter terrain update capability of the Viggen as a sort of analogy in my head.

[Swift.]

1 hour ago, frostycab said:

Regarding the Doppler, this could well be down to the fact that you were sat still and didn't go anywhere. I'm not familiar with exactly how it works on the Apache, but the name suggests the system depends on the principle of Doppler Shift, which is only observable when in motion relative to the signal. Presumably it uses a similar principle to the Rad Alt, and examines the frequency of the returned radio signals to derive the motion on the emitter, passing that onto the INUs to help correct them. When sat still on the ground like you are in the video the emitter would be getting no frequency change and thus has nothing to compare it's position to. Perhaps that might means it should tell the INUs that there's no motion so they stay aligned and centered, but there could be a weight-on-wheels lockout to protect ground crew.

I'd love to hear from someone who knows the system well and can explain. I've been using the radar altimeter terrain update capability of the Viggen as a sort of analogy in my head.

As I said, the doppler comment was related to the experience from the several hour long flight. Obviously in the video shown doppler isn't involved. However airborne or not, the INU continues to drift, so even if doppler *should* fix the INU, it doesnt in DCS.

[BIGNEWY]

I have marked reported, we have one open for INU drift currently 

thanks

[kaffepanna]

Hi.

I'm by no means smooth on the controls and good at keeping the Apache stable in a hover but I started to notice it was harder and harder the longer I have been flying. So after been practicing some hovering and doing touch and goes for a while, mostly starring at the symbology i look out the window and while the symbology says I'm perfectly centered and doing 0kn I'm basically at take of speed for a small aircraft. After landing and at a stand still the velocity vector is way back while standing still on the ground.

There is a big drift with the velocity vector in hover mode that get bigger over time. I have attached a trackfile and shows the drift after just a short while.

Is this a bug or should the aircraft really drift off this bad in such a short time?

 

vv-drift.zip

[Solo_Turk]

check this topic

[kaffepanna]

19 minutes ago, Solo_Turk said:

check this topic

I don't this is the same thing. The hover/bobup symbology actually shows a velocity in some direction when the aircraft is actually at a stand still. That doesnt go away when resetting the SAS

 

Edit added screenshot. And this is just after a few min of flying 

Edited April 13, 2022 by kaffepanna
added screnshot

[frostycab]

Degradation of the INU accuracy? Maybe try a reset and see what happens after a few minutes.

[Solo_Turk]

3 hours ago, kaffepanna said:

mostly starring at the symbology i look out the window and while the symbology says I'm perfectly centered and doing 0kn I'm basically at take of speed for a small aircraft

for this section. i believe you missed a feature. in hover mode if you fly fast and acceleration cue(little circle) reaches to ihads boundaries, it returns to the center of helmet cross but velocity line stays. so it doesn't mean you are flying 0kn with perfect hover. you are flying far beyond hover

Edited April 13, 2022 by Solo_Turk

[kaffepanna]

8 minutes ago, Solo_Turk said:

for this section. i believe you missed a feature. in hover mode if you fly fast and acceleration cue(little circle) reaches to ihads boundaries, it returns to the center of helmet cross but velocity line stays.

No I know that. thats not it. Its like in the screenshot (sitting still on ground) but even more off. (might have exaggerated the speed a bit)

 

13 minutes ago, frostycab said:

Degradation of the INU accuracy? Maybe try a reset and see what happens after a few minutes.

Im gonna try that. Still would it really work that way? For example helicopter thinks its at 5kn moving to the left when sitting still but as i can see the absolute position is correct

 

Edited April 13, 2022 by kaffepanna
clarification

[Crptalk]

 

[NineLine]

6 hours ago, kaffepanna said:

Hi.

I'm by no means smooth on the controls and good at keeping the Apache stable in a hover but I started to notice it was harder and harder the longer I have been flying. So after been practicing some hovering and doing touch and goes for a while, mostly starring at the symbology i look out the window and while the symbology says I'm perfectly centered and doing 0kn I'm basically at take of speed for a small aircraft. After landing and at a stand still the velocity vector is way back while standing still on the ground.

There is a big drift with the velocity vector in hover mode that get bigger over time. I have attached a trackfile and shows the drift after just a short while.

Is this a bug or should the aircraft really drift off this bad in such a short time?

 

vv-drift.zip 4.9 MB · 2 downloads

I dont see anything too out of the ordinary in your track, the velocity vector will look and act different in different modes. 

[kaffepanna]

6 hours ago, Crptalk said:

 

Thanks. Didn't find that post

6 hours ago, NineLine said:

I dont see anything too out of the ordinary in your track, the velocity vector will look and act different in different modes. 

That it looks different ant acts different in different modes is a given. The question here was about the drift and if it is to be expected. And as linked by Crptalk it has been seen before

[Swift.]

I think this is another report of the issue that everyone is seeing here. Uncorrectable INU velocity errors.

 

 

[BIGNEWY]

Posts merged. 

This is reported and the team will take a look 

thanks