Jump to content

LNAV Behaviour, obvious cause ?!


Recommended Posts

38 minutes ago, DGH said:

so “dude” where in your ads data do you see 90 degree turns  like on you initial picture ??

 

It's a fair question.

 

We would need to see the time stamp of the data.  While ADS-B can output every second (not taking into account inherent system latency onboard the aircraft), I doubt that Flight Aware is capable to providing 1 second position updates on aircraft - I wouldn't think their system simply can not handle that for the hundreds or few thousand aircraft they report positions on.  If this is the case, then that's why we sometimes see more acute angles reported.  I wasn't able to dig up flight awares positional update frequency, but I'm fairly certain I'm correct about this.

 

I hope this is at least a little helpful.

 

My very best wishes.

 

Link to comment
Share on other sites

2 hours ago, DGH said:

@LesOReilly

 

so “dude” where in your ads data do you see 90 degree turns  like on you initial picture ??

ADS out or not …. How come your initial picture is showing hard 90 degree UFO turns ???

 

your ads show a FEW degrees movement and not even a real back and forth wobble  … that is something else then what we are describing

 

 Sure the simulator is bugged …. Sure the crj is wobbling …. And yes …. There are also planes that do not show this wobble in that same simulator that is bugged … 

 

anyway lets end this.  Going nowhere. Have a nice weekend

Sorry you don't get to claim character issue and then "end it"...

 

You start off to claim that I am not showing the data with the picture... its the same data there is no 90 Degree turn.... then you say UFO turns??  what in the image shows that?  like what are you going on about?

 

You mean a portion of the flight path where it starts to turn to 33x direction... from it's previous heading?  they later in the flight when it turns more north on the next leg??

 

The point of the image was the middle section where the CRJ is moving back and forth to both sides....

 

And it is ADS-B Data not ADS....  You call my character into question and claim that I am fabricating something ... Yeah real stand up of you..

 

The Aircraft and Flight number is right there go look it up....  The Graph data is from it ... you can look it up yourself...  ... and that is an example of how the heading is fluctuating back and forth to both sides of the desired track... the point was to show that on paper and real life they are not "perfect" ...

 

Funny how again over in the WT Discord we have a conversation about people having issues with the CJ4 and the NXi having some GPS/LNAV tracking issues and wobbling back and forth and sometimes having to swerve a bit as they do intercepts and may over correct...

 

But yes we want it to be as best it can be... just need to be realistic on what a 50 dollar add on may or may not be able to do in an early platform that has had numerous growing pains and bugs..

 

 

 

Link to comment
Share on other sites

2 hours ago, DaveCT2003 said:

 

It's a fair question.

 

We would need to see the time stamp of the data.  While ADS-B can output every second (not taking into account inherent system latency onboard the aircraft), I doubt that Flight Aware is capable to providing 1 second position updates on aircraft - I wouldn't think their system simply can not handle that for the hundreds or few thousand aircraft they report positions on.  If this is the case, then that's why we sometimes see more acute angles reported.  I wasn't able to dig up flight awares positional update frequency, but I'm fairly certain I'm correct about this.

 

I hope this is at least a little helpful.

 

My very best wishes.

 

There is no 90 degree turns...  There are turns at the edges of the images... that had nothing to do with the conversation of it drifting and hunting along a course in the real world....

 

I did not claim that...  so no it is not a fair question ... it is trying to ignore the data point being that the plane is sending out it's Heading Data as part of the ADS-B...  So in the Middle Leg of those 3 Legs... you see that the plane goes left and right of the center track....  and the Data points is there to show that.

 

ADS-B updates are Done 1 per second.  Flight aware data posts are anywhere from 6-30 seconds depending on length of flight and the need to update the changes...

 

Again I will point out that the purpose was to show that as it is flying "down" that line the  heading is changing and the positions reports are to the Left and right of track.....  So yes the point is that the Real life CRJ is flying an LNAV course and it is not "spot on" straight lines...

 

 

 

 

 

 

Link to comment
Share on other sites

4 hours ago, LesOReilly said:

The Aircraft and Flight number is right there go look it up....

 

It actually is nowhere to be found in your screenshot or post...

 

I'm not quite sure what your point was posting that image. It shows a few degrees of change, which can be something as simple as a wind shift. But the issues discussed before are about completely incorrect turn anticipation, which is a whole different thing.

 

Ask any real CRJ pilot (I know you don't believe that I am one :P, but there are a few others around on here)... it does not do the crazy stuff that the sim addon does, nor is it constantly wobbling. And I know the claim of sim issues will come up again, and in return I'd like to point out that turn anticipation and weird LNAV stuff has been a constant thorn in the AS CRJs side since the CRJ X. 

 

Link to comment
Share on other sites

8 hours ago, LesOReilly said:

Sorry you don't get to claim character issue and then "end it"...

 

You start off to claim that I am not showing the data with the picture... its the same data there is no 90 Degree turn.... then you say UFO turns??  what in the image shows that?  like what are you going on about?

 

You mean a portion of the flight path where it starts to turn to 33x direction... from it's previous heading?  they later in the flight when it turns more north on the next leg??

 

The point of the image was the middle section where the CRJ is moving back and forth to both sides....

 

And it is ADS-B Data not ADS....  You call my character into question and claim that I am fabricating something ... Yeah real stand up of you..

 

The Aircraft and Flight number is right there go look it up....  The Graph data is from it ... you can look it up yourself...  ... and that is an example of how the heading is fluctuating back and forth to both sides of the desired track... the point was to show that on paper and real life they are not "perfect" ...

 

Funny how again over in the WT Discord we have a conversation about people having issues with the CJ4 and the NXi having some GPS/LNAV tracking issues and wobbling back and forth and sometimes having to swerve a bit as they do intercepts and may over correct...

 

But yes we want it to be as best it can be... just need to be realistic on what a 50 dollar add on may or may not be able to do in an early platform that has had numerous growing pains and bugs..

 

 

 


well “dude” 

 

i think the perfectly smooth GREEN LINE in you initial picture is the CRJ you selected … that line shows NO wobble as addressed by people on the forum.

 

you see. Big difference between the “wobble” in that green line and the examples shown on the forum.

 

if you can not see that there is no point is discussing 

 

people here are talking about full ( halve actually) bank left …. Then right … then left again. NOT a few degrees of wind correction / drift !!! A few degrees is not an issue for anyone !

 

do you know an airway is 8 miles wide ? Planes move SLOWLY in those airways to compensate drift and wind. Airplanes do not follow a flightaware dotted line perfectly.  That dotted line is just a line between A and B. That line does not calculate wind gust and turbulence !!!
 

No airliner will bounce from left to right in airways like an aggressive ping pong ball ( UFO 😬) !! THAT is what people complaint about.

 

read the forum and check the differences between your screenshot and the FS2020 screenshots. We are talking on complete different “wobbling”.

 

for your ease … i zoomed a little and added the screenshots below.

 

you talk on : SLOW few degree wind correction / drifting passengers wil not even notice … in your picture.

 

We talk on : violent left right left pingponging at max available bank angle (1/2 at cruising) … compensating just to make a turn !!! 
 

If you do not see this and try to proof “ real wold pax are thrown left and right just like the FS2020 CRJ” …. By showing a few degrees of drift ADS drift data ( one direction not even a wobble !! ) …. There is no further point in discussing 

 

i agree my previous post where a little unclear. Tried some sarcasm jokes but that is not working here .. sorry for that and hope this post was more clear.

 

so with this i will bug out the discussion from my side 👍

 

you

Please login to display this image.


we

Please login to display this image.

  • Like 1
Link to comment
Share on other sites

14 hours ago, LesOReilly said:

There is no 90 degree turns...  There are turns at the edges of the images... that had nothing to do with the conversation of it drifting and hunting along a course in the real world....

 

I did not claim that...  so no it is not a fair question ... it is trying to ignore the data point being that the plane is sending out it's Heading Data as part of the ADS-B...  So in the Middle Leg of those 3 Legs... you see that the plane goes left and right of the center track....  and the Data points is there to show that.

 

ADS-B updates are Done 1 per second.  Flight aware data posts are anywhere from 6-30 seconds depending on length of flight and the need to update the changes...

 

Again I will point out that the purpose was to show that as it is flying "down" that line the  heading is changing and the positions reports are to the Left and right of track.....  So yes the point is that the Real life CRJ is flying an LNAV course and it is not "spot on" straight lines...

 

 

 

 

 

 

 

We have a wide variety of people here from different backgrounds and knowledge levels.  What is important is to remain educational in our posts without taking things personally or feeling challenged by someone, and even if we are a well meaning, educational reponse may difuse things and return everything back to a status quo. It doesn't always work, but it works enought to keep things educational and on an even keel.

 

I think hope you saw my explanation of what the person was seeing (graphically there was indeed what APPEARED to be a 90 degree turn).  The rest of what you posted above was my explaination.  I wasn't as specific as I could have been because I wanted to avoid the appearance of acting above the other person.  Just an easy explaination that hopefully made sense to him and in support of what you had said.  Why you directed you post back at me has me a bit mystified, but not bothered by it.

 

My best wishes to you.

 

  • Thanks 1
Link to comment
Share on other sites

On 9/4/2021 at 2:57 AM, CRJay said:

 

It actually is nowhere to be found in your screenshot or post...

 

I'm not quite sure what your point was posting that image. It shows a few degrees of change, which can be something as simple as a wind shift. But the issues discussed before are about completely incorrect turn anticipation, which is a whole different thing.

 

Ask any real CRJ pilot (I know you don't believe that I am one :P, but there are a few others around on here)... it does not do the crazy stuff that the sim addon does, nor is it constantly wobbling. And I know the claim of sim issues will come up again, and in return I'd like to point out that turn anticipation and weird LNAV stuff has been a constant thorn in the AS CRJs side since the CRJ X. 

 

 

I actually never said I don't believe you are a real CRJ Pilot ....  I said if I need to get clarity on something I can ask my buddies that I know who ARE CRJ Pilots... As in people I know... But what ever you want to bring up posts from the MSFS forum?  we can go down that path if you like..

 

I think what my point about it was is this --- "it's never perfect" can they do better yes... but I find it strange that we have similar oscillation issues in other MSFS Planes including the new NXi that is complete from the ground up custom and drives the "SIM" in Pitch and Roll mode.....  yet people are not thrashing on WT...  Same with the CJ4 it can have some weird moments... again people don't thrash on them...

 

Flight I looked up is JAZZ 8889

 

Point was that unlike the manuals that always show a perfect line that fact is the planes do oscillate.....  Now can they do a better and try to make it a better experience .. probably...  But do they really need people piling stuff on?  

  • Like 1
Link to comment
Share on other sites

On 9/4/2021 at 3:32 AM, DGH said:

 

 

do you know an airway is 8 miles wide ? Planes move SLOWLY in those airways to compensate drift and wind. Airplanes do not follow a flightaware dotted line perfectly.  That dotted line is just a line between A and B. That line does not calculate wind gust and turbulence !!!

 

Okay I will only reply to this part ... 

 

The dotted line is the filed plan route... and yes that is the center of the airway or path that the plane is programmed to fly....  I am fully aware how wide and airway is... I have flown them many times...  And you are proving my point that wind and other factors cause the plane to not be able to keep a perfect heading the entire time... that it is not able to calculate are react in a manner that allows it to stay perfectly locked on TRACK...  That section is literally the Q969 airway.... which flight aware draws and so the point I was making was to your reference that the manuals show perfect lines and I just wanted to point out that it's never really a perfect line....  

 

What is your endgame in all of this?  It's been pointed out that improvement can be made.. 

 

 

Link to comment
Share on other sites

3 hours ago, LesOReilly said:

I think what my point about it was is this --- "it's never perfect" can they do better yes...

 

Your example is still irrelevant if you look at the issues mentioned here, which are mainly about turn anticipation not being implemented properly. Has nothing to do with some slight heading variations when correcting for wind change or being 'locked on' to an airway track, these are pure and simple incorrect turns.

  • Upvote 3
Link to comment
Share on other sites

Dear Devs,

 

since 24. August this year ;-),  not a single comment from you guys on this topic.

Valuable members of this forum have chimed in and i am very pleased with their comments and suggestions. Thank you all for that.
Just saying

 

Oliver

Link to comment
Share on other sites

On 9/9/2021 at 5:16 AM, JetNoise said:

Dear Devs,

 

since 24. August this year ;-),  not a single comment from you guys on this topic.

Valuable members of this forum have chimed in and i am very pleased with their comments and suggestions. Thank you all for that.
Just saying

 

Oliver

The developers are definitely aware of this problem, and a new system for intercepting the outbound leg when the course changes at a waypoint is being implemented, based on how the real CRJ autopilot operates.
 

The initial part of a course change generally works exactly as it would in the real CRJ. The problem (at present) is with the last part of the course change. The actual autopilot is limited to two fixed bank angles when turning under autopilot control - either 25 or 12.5 degrees. (The sim uses 30 and 15 degrees).

 

Above 31,600 feet, the only available bank angle is 12.5 degrees. For this reason, the real CRJ will use a longer DTA (turn anticipation distance) than other types of airliners with a more sophisticated autopilot that can dynamically vary the bank angle in a turn.  
 

When the autopilot is in half bank mode, the real CRJ will use a DTA of up to 11 miles prior to the waypoint if the course change is significant.
 

The calculation of turn anticipation is based on the airspeed, the available bank angle, and the amount of heading change required. With a bank angle of 12.5 degrees at high altitude, even a DTA of 11 miles may not be enough with a significant tailwind, and the real aircraft can overshoot the new course line in this scenario.

 

The main issue with the sim version at present is not overshoot but undershoot - i.e. the aircraft completes the turn before arriving at the new course, and then makes a series of “cut and try” banks to work its way onto the new course.

 

The actual autopilot switches from constant bank mode to FMS CDI tracking during the final part of an LNAV turn. If the aircraft has not yet arrived at the new course, the aircraft will roll level, and monitor the FMS CDI deflection and centering rate to control the point where it will make the final bank to roll out on course. This is similar to how the autopilot intercepts a localizer. 
 

This change being incorporated into the sim CRJ, should result in a significant improvement in intercepting a new course in LNAV mode, without the constant banking seen now. Even the real CRJ may have to make one or two corrective banks once established on a new course to acquire the correct wind correction angle, but that is different than the long series of turns that currently can happen.

  • Like 2
  • Thanks 3
  • Upvote 2
Link to comment
Share on other sites

vor 6 Stunden , amahran sagte:

Thanks, @JRBarrett!

 

I was constantly peeved that the CRJ was taking these weird turns, thinking I was screwing up somewhere. But this post was actually insightful in helping me feel I'm not alone!

 

I can only second that @JRBarrett.   
Very comprehensive and insightful and very much appreciated !!!
 

... I had a flight yesterday where it took 7 (seven) turns to get back on track (~80kts of winds involved, FL360) .

On the other hand i have seen a departure with an almost 180 turn (as per SID), flown almost perfect (imho) ...
 

Oliver

Link to comment
Share on other sites

On 9/15/2021 at 3:09 AM, JetNoise said:

 

I can only second that @JRBarrett.   
Very comprehensive and insightful and very much appreciated !!!
 

... I had a flight yesterday where it took 7 (seven) turns to get back on track (~80kts of winds involved, FL360) .

On the other hand i have seen a departure with an almost 180 turn (as per SID), flown almost perfect (imho) ...
 

Oliver

The excessive turning after a course change in the current version is by no means a “given”. Most of the time it works well. But, there are times it does not capture decisively.
 

The worst case scenario seems to be when there is a substantial course change at high altitude and airspeed, with a strong crosswind coming from the outside of the turn, as this tends to push the aircraft away from the new track as it is trying to capture. 
 

Any changes to be made in this area will be thoroughly tested before being released. 

  • Thanks 1
Link to comment
Share on other sites

G-Day,

 

today another strange occurance of LNAV not behaving

Please login to display this image.

 

A/C  inbound KOR waypoint, started its turn towards PIKAD waypoint. It continued on that left turn WITHOUT correcting to PIKAD.

 

I did Heading correction (inbound track and a Paste & Copy IXONI waypoint .  Toggled NAV and A/C continued as expected. 
Winds 294/35  

 

Oliver

Link to comment
Share on other sites

Hi again,

 

half way through my flight, FMC untouched, i see this 

 

Please login to display this image.

 

Some Magic is happening ...

 

Could not get it fixed either:  Deleting RSKs , Resetting Altitude in PERF, CLMB/DESC via AP and Level OFF ...

Oliver

Link to comment
Share on other sites

On 16.9.2021 at 16:23, JRBarrett sagte:

The worst case scenario seems to be when there is a substantial course change at high altitude and airspeed, with a strong crosswind coming from the outside of the turn, as this tends to push the aircraft away from the new track as it is trying to capture. 

I can confirm this. That scenario is indeed the most problematic one.

 

On 18.9.2021 at 17:10, JetNoise sagte:

half way through my flight, FMC untouched, i see this 

 

Please login to display this image.

 

 

Unrelated to the LNAV behavior, but anyway: Looks like there is a "below 8000" constraint at ROBEG:

Please login to display this image.

 

Maybe the FMC got confused by that. The full route would be interesting.

Link to comment
Share on other sites

vor 14 Stunden , SinusJayCee sagte:

I can confirm this. That scenario is indeed the most problematic one.

 

 

Unrelated to the LNAV behavior, but anyway: Looks like there is a "below 8000" constraint at ROBEG:

Please login to display this image.

 

Maybe the FMC got confused by that. The full route would be interesting.

 

G'morning 😉

 

Route:  

LGAV/21L KOR1F KOR UL53 GARTA UV60 YNN UL604 DIMIS DCT RETRA DCT
PERAN DCT IRDIV DCT PETOV DCT LAMSI DCT LASGA M726 LASTO M852 BIRKA
T803 GITEX GITEX4R EDDV/09L

 

ROBEG (IAF) has no constraint. Next DESIM, 3000ft is required to catch the ILS into 9L EDDV.

Wenn i initially checked the FPL (SID,STAR,APP), all was okay. It happened inflight, magically, the FMC shows still 577NM away from DEST.

 

 

 

Happy days

Oliver

 

 

Link to comment
Share on other sites

Vor 1 Stunde, JetNoise sagte:

ROBEG (IAF) has no constraint. Next DESIM, 3000ft is required to catch the ILS into 9L EDDV.

GITE4R has indeed no constraint at ROBEG, but the ROBE09 RNAV transition has. Did you accidentally select this transition or was it automatically selected? Iirc, the CRJ sometimes preselects transitions for approach.

 

Vor 1 Stunde, JetNoise sagte:

Wenn i initially checked the FPL (SID,STAR,APP), all was okay. It happened inflight, magically, the FMC shows still 577NM away from DEST.

The selected transition may also explain the strange distance value, because itcontains a vector segment, which may have confused the FMS. It has already been reported that there are some issues with vector segments. However, it obviously shouldn't display those large distances and this doesn't explain why it happened suddenly.

Link to comment
Share on other sites

vor 3 Stunden , SinusJayCee sagte:

GITE4R has indeed no constraint at ROBEG, but the ROBE09 RNAV transition has. Did you accidentally select this transition or was it automatically selected? Iirc, the CRJ sometimes preselects transitions for approach.

 

The selected transition may also explain the strange distance value, because itcontains a vector segment, which may have confused the FMS. It has already been reported that there are some issues with vector segments. However, it obviously shouldn't display those large distances and this doesn't explain why it happened suddenly.

Quick: I didn't use the transition....

I am aware of vector segments and their "problems"

 

Oliver

Link to comment
Share on other sites

vor 40 Minuten, JetNoise sagte:

Quick: I didn't use the transition....

My point was that it could have been selected automatically without you actively choosing it. Or you accidentally selected it by choosing "via ROBEG", because the standard ILS 09L procedure starts at ROBEG. Otherwise I cannot explain the below 8000 constraint at ROBEG. Other cause may be that there is something wrong in the nav data and it contains a constraint at the end of GITE4R.

Link to comment
Share on other sites

Vor 1 Stunde, SinusJayCee sagte:

My point was that it could have been selected automatically without you actively choosing it. Or you accidentally selected it by choosing "via ROBEG", because the standard ILS 09L procedure starts at ROBEG. Otherwise I cannot explain the below 8000 constraint at ROBEG. Other cause may be that there is something wrong in the nav data and it contains a constraint at the end of GITE4R.

Well, you may be right on this, but that does not explain the FL340 at BIRKA as you can see on my screenshot of the FMC above....

We will have to wait for the next update and see if things have been fixed.

The CRJ is definitely not one of all my other aircraft you can leave alone for a p....

(732,733,Maddog, DC-6 ...)

Oliver

Link to comment
Share on other sites

vor 35 Minuten, JetNoise sagte:

Well, you may be right on this, but that does not explain the FL340 at BIRKA as you can see on my screenshot of the FMC above....

That's not what I was trying to say. I just tried to figure out where the 8000 came from. That the FMS got confused here somehow is not a question.

 

vor 36 Minuten, JetNoise sagte:

We will have to wait for the next update and see if things have been fixed.

Unless there is a reliably way to reproduce the issue, the devs cannot check it and accordingly cannot fix it in the next update.

Link to comment
Share on other sites

  • 2 weeks later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. Privacy Policy & Terms of Use