reported INU drifts / INU fix problems still?

[Usagi]

I am having a lot of problems with navigation systems of Ka-50 BS3 (latest open beta MT 2.9.0.47168).

INU begins to drift right after takeoff, often losing heading completely. That makes use of datalink and marking up targets useless as all markings will be way off the real location.

 

 

Here I have flown just 02:33 and navigation system thinks that I should go 5 degrees left from real heading.

Got also 'missile 3 o'clock low' warning from EKRAN when INU FIX error appeared.

 

Same mission, second attempt. Here navigation system would guide me 10 degree to the right.

Edited October 28, 2023 by Usagi
lost all text after first post, rewrote

[Hobel]

vor 16 Stunden schrieb Usagi:

I am having a lot of problems with navigation systems of Ka-50 BS3 (latest open beta MT 2.9.0.47168).

INU begins to drift right after takeoff, often losing heading completely. That makes use of datalink and marking up targets useless as all markings will be way off the real location.

 

 

Here I have flown just 02:33 and navigation system thinks that I should go 5 degrees left from real heading.

Got also 'missile 3 o'clock low' warning from EKRAN when INU FIX error appeared.

 

Same mission, second attempt. Here navigation system would guide me 10 degree to the right.

 

Did you also observe this before 2.9?  

 

What alignment are you running?  Fast alignment,   Norm or  Precise?

 

 

[Usagi]

3 hours ago, Hobel said:

Did you also observe this before 2.9?  

 

What alignment are you running?  Fast alignment,   Norm or  Precise?

 

 

I use normally Fast alignment. If mission starts with engines on, its usually Normal + gyro in mission builder.

Just tested with stable release and it's all the same.

Just after couple of minutes flight I get 'missile 3 o'clock low' warning and Inu begins to have more and more error.

[Hobel]

vor 6 Stunden schrieb Usagi:

I use normally Fast alignment. If mission starts with engines on, its usually Normal + gyro in mission builder.

Just tested with stable release and it's all the same.

Just after couple of minutes flight I get 'missile 3 o'clock low' warning and Inu begins to have more and more error.

Has this always been the case or has the behavior changed after 2.9?  

  

vor 6 Stunden schrieb Usagi:

Just after couple of minutes flight I get 'missile 3 o'clock low' warning and Inu begins to have more and more error.

Are you suggesting that the missle warning provocates a faster INU drift?

and btw
Fast Aligment is not intended for combat use, therefore strong drifts are to be expected after a short time, especially with stronger maneuvers over +15 bank.
herefore Norm or Prec is necessary especially if you want to use Datalink and similar precisely.



nevertheless I will have another look at it, you never know, otherwise you can upload a track of your situation.

Edited October 29, 2023 by Hobel

[Usagi]

Hi,

It seems that Normal and Precision alignment does the trick. Still not sure how hot start is.

The problem lies in many (campaign) missions that has time limits. You cannot perform Normal INU alignment as it takes too much time.

"Easy Start" also does only fast alignment.

[Hobel]

vor 5 Stunden schrieb Usagi:

The problem lies in many (campaign) missions that has time limits. You cannot perform Normal INU alignment as it takes too much time.

okay this is an interesting problem, theoretically it could be fixed if the mission builders change the start conditions for the KA-50III in the editor.
You don't have access to it yourself?

  

vor 5 Stunden schrieb Usagi:

Still not sure how hot start is.

there is a selection for this in the mission editor

@Flappie can you take a look at it.

Edited November 6, 2023 by Hobel

[Flappie]

@Hobel Sure.

@Usagi Please name at least one of the campaigns you're facing the problem with. I'll check.

[Usagi]

On 11/7/2023 at 7:19 PM, Flappie said:

@Hobel Sure.

@Usagi Please name at least one of the campaigns you're facing the problem with. I'll check.

 

Hi,

Pandemic Campaign and Republic Campaign

[Flappie]

I've just checked three Ka-50 campaigns:

• Pandemic has the player Ka-50 set to "No alignment and fixtaking needed".
• MAD has the player Ka-50 set to "No alignment and fixtaking needed".
• Republic has the player Ka-50 set to "Fully realistic" (alignment + fixtaking needed).

Do you remember which Repulic mission(s) have a time limit?

EDIT: I think I found one, Ambush ("Fly towards objective "Ambush" to 12:15, hover and wait for the Mi-24’s. ETA of the Mi-24s - 12:20").

I've just noticed that a player Ka-50 III added to the ME is automatically set to "Fully realistic", so I think this is an error. Issue reported for the Republic campaign.

Edited November 18, 2023 by Flappie

[PawlaczGMD]

7 hours ago, Flappie said:

I've just checked three Ka-50 campaigns:

• Pandemic has the player Ka-50 set to "No alignment and fixtaking needed".
• MAD has the player Ka-50 set to "No alignment and fixtaking needed".
• Republic has the player Ka-50 set to "Fully realistic" (alignment + fixtaking needed).

Do you remember which Repulic mission(s) have a time limit?

EDIT: I think I found one, Ambush ("Fly towards objective "Ambush" to 12:15, hover and wait for the Mi-24’s. ETA of the Mi-24s - 12:20").

I've just noticed that a player Ka-50 III added to the ME is automatically set to "Fully realistic", so I think this is an error. Issue reported for the Republic campaign.

 

I've played the Pandemic a few months ago, and all the target points were wayyy off when slewed to... This might have changed since.

[Hobel]

vor 11 Minuten schrieb PawlaczGMD:

I've played the Pandemic a few months ago, and all the target points were wayyy off when slewed to... This might have changed since.

if you see it again send me a track i'll take a look at it

[PawlaczGMD]

4 minutes ago, Hobel said:

if you see it again send me a track i'll take a look at it

I'll try to fly a mission some time soon. But it has worked the same way as Republic for me - target points and datalink points are useless. The briefing lacks info about designated fix points. The INU drifts a lot anyway because full alignment wasn't done.

[Hobel]

vor 9 Stunden schrieb PawlaczGMD:

I'll try to fly a mission some time soon. But it has worked the same way as Republic for me - target points and datalink points are useless. The briefing lacks info about designated fix points. The INU drifts a lot anyway because full alignment wasn't done.

as I said, show me the track then i  can rule out user errors
 

Zitat

The briefing lacks info about designated fix

this is not necessary you can either create them yourself or you can use the information from the Abriss to fix the INU.

Here is an example of how it works with the Shark Planner (<Link), these things can also be entered manually, but of course it is faster this way, unfortunately DCS still lacks such tool.
In the video I cause a 1km drift.
and then I create a fix point ingame and use it directly, as you can see it's no problem

 

 

and here is an example, you can enter the coordinates in the PVI-800 and then it is up-to-date again

 

Edited November 19, 2023 by Hobel

[PawlaczGMD]

50 minutes ago, Hobel said:

as I said, show me the track then i  can rule out user errors
 

this is not necessary you can either create them yourself or you can use the information from the Abriss to fix the INU.

Here is an example of how it works with the Shark Planner (<Link), these things can also be entered manually, but of course it is faster this way, unfortunately DCS still lacks such tool.
In the video I cause a 1km drift.
and then I create a fix point ingame and use it directly, as you can see it's no problem

 

 

and here is an example, you can enter the coordinates in the PVI-800 and then it is up-to-date again

 

 

Thanks for the videos. I'll save a track when I encounter something that feels off.

I just think that if a mission implements a fix point, it should tell you what it is. The problem I've noticed is that without precise alignment, The INU drifts fast after doing a fix anyway, so it's not really worth doing it manually. This software is pretty nice.

What are you actually doing in the second video?

BTW, the ABRIS has your coordinates from GPS, right? So can you just do a fly-over fix on yourself from ABRIS coords? 

[Hobel]

vor 3 Minuten schrieb PawlaczGMD:

BTW, the ABRIS has your coordinates from GPS, right? So can you just do a fly-over fix on yourself from ABRIS coords? 

yes, this is the point of the second video you can enter the displayed coordinates into the PVI-800.

In the past, the ABriss was supposed to update the PVI-800 automatically every 300 seconds, but unfortunately ED abandoned this plan ¯\_(ツ)_/¯

 

[PawlaczGMD]

7 minutes ago, Hobel said:

yes, this is the point of the second video you can enter the displayed coordinates into the PVI-800.

In the past, the ABriss was supposed to update the PVI-800 automatically every 300 seconds, but unfortunately ED abandoned this plan ¯\_(ツ)_/¯

 

It is very unreasonable that it doesn't work that way. Because ABRIS and INU have a way of talking to each other - Datalink points are saved/displayed to ABRIS through the INU, and Shkval target is displayed on the ABRIS if I understand it correctly (Because only the INU knows directly where you're pointing the Shkval, right?). So to not have it send GPS coords every X seconds is weird. Maybe ABRIS is read-only?

[Hobel]

vor 7 Minuten schrieb PawlaczGMD:

It is very unreasonable that it doesn't work that way. Because ABRIS and INU have a way of talking to each other - Datalink points are saved/displayed to ABRIS through the INU, and Shkval target is displayed on the ABRIS if I understand it correctly (Because only the INU knows directly where you're pointing the Shkval, right?). So to not have it send GPS coords every X seconds is weird. Maybe ABRIS is read-only?

According to the ED statement, the ABRIS cannot send such data to the PVI-800, so the plan was rejected.

I would also think that at least the KA-50III should be able to do this because the KA-50III is an ED product with the "what if?" philosophy.

[PawlaczGMD]

5 minutes ago, Hobel said:

According to the ED statement, the ABRIS cannot send such data to the PVI-800, so the plan was rejected.

I would also think that at least the KA-50III should be able to do this because the KA-50III is an ED product with the "what if?" philosophy.

Well, it's a dodgy area, because KA-50 is a prototype anyway, with individual airframes having different mods. I think the Igla capacity comes from just a mockup. So they could have gone either way with this. It is really annoying in practice if the mission doesn't start you aligned.

[Hobel]

vor 8 Stunden schrieb PawlaczGMD:

It is very unreasonable that it doesn't work that way. Because ABRIS and INU have a way of talking to each other - Datalink points are saved/displayed to ABRIS through the INU, and Shkval target is displayed on the ABRIS if I understand it correctly (Because only the INU knows directly where you're pointing the Shkval, right?). So to not have it send GPS coords every X seconds is weird. Maybe ABRIS is read-only?

Yes, you can enter tgt in the INU, which will then be displayed in the Abris.   I can't remember exactly why it works this way but not the other way around. @okopanjaCould explain this aspect better

 

 And as a reminder, fast alignment is not intended for combat use, especially if you want to work with target points and the INU

 

 

[okopanja]

3 hours ago, Hobel said:

Yes, you can enter tgt in the INU, which will then be displayed in the Abris.   I can't remember exactly why it works this way but not the other way around. @okopanjaCould explain this aspect better

@PawlaczGMDFrom what I gathered so far and found in internet. Ka-50 at least 2 major versions of cockpit one without ABRIS and one with ABRIS. Additional sub-variants are possible including different SHKVAL screen (e.g. green vs brown). The main system navigation/attack system is called Rubikon K-041 and interacts with pretty much everything. The navigation part of the system in particular is called PNK-800 Radian and consist of several devices including INU. When entering coordinates through PVI-800 you are actually entering them into PNK-800 and Rubikon. Rubikon utilizes SHKVAL to perform the Fixpoint-based update.

It can be assumed that Rubikon predates integration of the ABRIS, and that ABRIS was added later, as relatively simply upgrade (it caused the cockpit layout change). There are indications that the ABRIS was actually a product of Swedish company that specializes in civilian navigation. In addition you will notice the Garmin antenna on the 3D model of Black shark.

So far the following information from Rubikon is displayed into ABRIS:

• navigation points
• fix points
• targets points
• azimuth and distance to the SHKVAL view point (this is sadly not accurate, but more later...)

You will notice that in mission editor these values are exactly spot on with ABRIS equivalents of these points. This is because mission editor on load loads them into Rubikon with absolute precision. If this is the case the targets will be spot on when you slew the shkval to target points and ready to fire.

Now following applies only to manual entry:

PVI-800 allows you the entry in (D)DD MM.m format which will introduce relatively large error for latitude. This is roughly +-92,5m. You can calculate this yourself when you know that 1 degree of latitude has certain number of meters (I used one of the online values, and did not check how much it is actually in DCS ).  The error of the longitude depends actually on the latitude (parallels are not equally sized) is typically lower than latitude errors.

You can not expect to slew the Shkval to the target and immediately lock.

In addition the Rubikon seems to be unaware of the altitude of target/navigation/fix points. Instead it, in DCS it will assume that point is at see level. As a consequence:

- target points will be usable only at up to 100m altitude

- SHKVAL indication in ABRIS will have usable indication of distance up to 100m altitude

- in all other altitudes targets points from Rubikon and SHKVAL indication in ABRIS will be almost useless

I did interact with ED developer and pointed out that:

1. targeting system that works only next to the see level is illogical, and that Rubikon must have some knowledge on e.g. target point altitude. He did propose that maybe they can assume the target altitude based on helicopter altitude above see. I proposed the algorithm based on the laser range finder of SHKVAL (basically starting with his proposal augmented with iteration along the azimuth until the coordinates roughly match). Still no feedback if they are going to do anything, but I gather they were busy with DLSS and damn spotting dots.

2. manual contains written text and pictures that indicate that PNK-800 has information on altitude of waypoints. I argued that there must be  a way to enter those values in PVI-800. It should be noted that there is even a flag in ADI that shows direction of deviation of altitude compared to current waypoint). This flag actually works if you started with preloaded coordinates.  So far I got the answer that manual contains errors.

At the end I attached the PNG and SVG diagram of devices generated from ED's own code with graphviz.

Once you download the SVG to your local storage, open it with web browser. SVG contains additional tooltip annotations.

ka-50.svg

[okopanja]

I hope you do not mind I sort of went out of the topic. I have not flown the Ka-50 in 2.9 that much recently, so I will also try to replicate the problem.

[PawlaczGMD]

2 hours ago, okopanja said:

@PawlaczGMDFrom what I gathered so far and found in internet. Ka-50 at least 2 major versions of cockpit one without ABRIS and one with ABRIS. Additional sub-variants are possible including different SHKVAL screen (e.g. green vs brown). The main system navigation/attack system is called Rubikon K-041 and interacts with pretty much everything. The navigation part of the system in particular is called PNK-800 Radian and consist of several devices including INU. When entering coordinates through PVI-800 you are actually entering them into PNK-800 and Rubikon. Rubikon utilizes SHKVAL to perform the Fixpoint-based update.

It can be assumed that Rubikon predates integration of the ABRIS, and that ABRIS was added later, as relatively simply upgrade (it caused the cockpit layout change). There are indications that the ABRIS was actually a product of Swedish company that specializes in civilian navigation. In addition you will notice the Garmin antenna on the 3D model of Black shark.

So far the following information from Rubikon is displayed into ABRIS:

• navigation points
• fix points
• targets points
• azimuth and distance to the SHKVAL view point (this is sadly not accurate, but more later...)

You will notice that in mission editor these values are exactly spot on with ABRIS equivalents of these points. This is because mission editor on load loads them into Rubikon with absolute precision. If this is the case the targets will be spot on when you slew the shkval to target points and ready to fire.

Now following applies only to manual entry:

PVI-800 allows you the entry in (D)DD MM.m format which will introduce relatively large error for latitude. This is roughly +-92,5m. You can calculate this yourself when you know that 1 degree of latitude has certain number of meters (I used one of the online values, and did not check how much it is actually in DCS ).  The error of the longitude depends actually on the latitude (parallels are not equally sized) is typically lower than latitude errors.

You can not expect to slew the Shkval to the target and immediately lock.

In addition the Rubikon seems to be unaware of the altitude of target/navigation/fix points. Instead it, in DCS it will assume that point is at see level. As a consequence:

- target points will be usable only at up to 100m altitude

- SHKVAL indication in ABRIS will have usable indication of distance up to 100m altitude

- in all other altitudes targets points from Rubikon and SHKVAL indication in ABRIS will be almost useless

I did interact with ED developer and pointed out that:

1. targeting system that works only next to the see level is illogical, and that Rubikon must have some knowledge on e.g. target point altitude. He did propose that maybe they can assume the target altitude based on helicopter altitude above see. I proposed the algorithm based on the laser range finder of SHKVAL (basically starting with his proposal augmented with iteration along the azimuth until the coordinates roughly match). Still no feedback if they are going to do anything, but I gather they were busy with DLSS and damn spotting dots.

2. manual contains written text and pictures that indicate that PNK-800 has information on altitude of waypoints. I argued that there must be  a way to enter those values in PVI-800. It should be noted that there is even a flag in ADI that shows direction of deviation of altitude compared to current waypoint). This flag actually works if you started with preloaded coordinates.  So far I got the answer that manual contains errors.

At the end I attached the PNG and SVG diagram of devices generated from ED's own code with graphviz.

Once you download the SVG to your local storage, open it with web browser. SVG contains additional tooltip annotations.

ka-50.svg 103.22 kB · 0 downloads

Thanks for all the detail! I was unaware of the altitude issue. I've assumed that it gets the target altitude from laser ranging. This might explain why points are useless in some conditions. 

[okopanja]

5 minutes ago, PawlaczGMD said:

Thanks for all the detail! I was unaware of the altitude issue. I've assumed that it gets the target altitude from laser ranging. This might explain why points are useless in some conditions. 

When you use the shkval for fix point update, the Rubikon does not know the altitude of the fix point. However, it knows fix point coordinates + own altitude and azimuth (there is electronic compass among the devices). When you point the shkval to fix point and perform the action it will record shkval azimuth, and elevation angle, while measuring direct distance (not horizontal), based on that it will calculate the coordinate update.

 

[PawlaczGMD]

I just started playing the MAD campaign, which starts with precise alignment and no drift, and everything related to navigation works without issues. 

[MarcT-NL]

I just bought the MAD campaign, but the INS is all over the place. I do a normal startup. At least I think I do.... I don't get flashing HUD or countdowns or so...