Richard Dastardly
-
Posts
186 -
Joined
Content Type
Profiles
Forums
Downloads
Posts posted by Richard Dastardly
-
-
On 3/23/2025 at 11:29 AM, Mathijs Kok said:
They need a simple tweak. See the attached document Hans shared with me.
Mathijs Kok
PMDG
Thanks Mathjis, no sign of model changes then ( thankfully, I'm not sure where the source for my liveries is atm... ).
---
I've never had a problem with it intercepting a LOC, it's never been able to fly a hold though ( and I've gone one at length about how comical it is, including track maps
). Was the released patch actually the right one? I mean, the ATR is quite capable of flying a hold...
-
1
-
1
-
-
Are there any model changes that'll need reworks of existing paints?
-
15 hours ago, Joe Markowski said:
Was this the issue where if you don't Exec after adding a runway or approach to the FMS it will double up on the STAR waypoints?
No, it was the "wandeirng all over the sky instead of attempting to follow the flightplan" problem
I think it was that approach to Stansted that sends you in a loop to have a look at Duxford, but it might have been a different and a bit more direct one.
7 hours ago, Syrinx said:I believe it's happening when I'm not going fast enough.
I had the VNAV cruise say .74M, and suggest it can fly x altitude at that speed with my loadout; it lied. When I pushed the throttles 'through the firewall' and got her up to .80M, the waddling stopped.I'm always at 0.81 or more if I can ( recently seems the -900 won't go any faster, no more 0.84m cruising! ) - I try not to fly actually on the stop below climb but it's usually pretty near. Still get some LNAV issues.
-
It's certainly not manual input doing it, because I never import routes. I've flown it recently but not since SU12 ( and I can't right now because my motherboard is dying ), it was *mostly* ok provided I remembered the stick profile with the deadzone, and remembered all the things you can't tell it to do. It made a real mess of one STAR, I remember.
-
@Mathijs KokIs it under your control whether you can unencrypt the config files for the marketplace release? the premium edition bundled aircraft have been decrypted but generally marketplace ones are still as they were. I too unfortunately bought both CRJ packages from the market before I knew any better...
-
This was covered in the "LNAV Tracking Issue" thread at some length, there appear to be two separate problems with holds - first off you can't program them right, the aircraft wants to fly a reciprocal of the heading you enter the hold at rather than what you tell it to - and then it won't fly the hold pattern either ( which may or may not be something to do with the other issue, might be something to do with the general LNAV issue, or might be something entirely new ). If you could provide tracks ( I take mine from LittleNavMap ) along with hold details & what you programmed in I'm sure it'd probably help when it comes to diagnosing the problem.
-
1
-
-
On 1/4/2023 at 1:51 PM, DGH said:
Luckily we have working 737 LNAV ... no back and fourth swaying and happy passengers.
Almost 2 year anniversary of this LNAV bug ...
ANY NEWS ?
Given how the DC-6 AP is currently going all over the sky too, I wouldn't go too far down that hole
I haven't actually flown any of the CRJs in months, I'm presuming SU10 & 11 didn't help at all?
-
1
-
-
Fairly sure Hans is working on the ATR for Asobo atm...
-
2
-
-
On 9/28/2022 at 6:43 PM, Tomcat_22 said:
Perhaps I phrased it wrong, I've not got the 1st clue of modding so I'm just blabbering if you will, getting rid of some frustrations I guess, what I meant with a modding is having the plance force read a different AP system or one that over-rides the default one, at least the LNAV, heck I don't know, but how it's currently, is absolutely not fun.
As for reusing some code, I believe their next plane is the A330, maybe, not sure, I should hope they wont use any of this code, probably can't anyways as the FBW has different logic... But yeah, at the end of the day, I hope this gets fixed because I've also parked the CRJ for now, I'll use a bigger GA plane or the 736 for short hops, one that can actually fly an LNAV plan.I've no doubt you could insert your own code in there if you patched the entire plane to look at it instead of the aerosoft code, but then you'd have to write an awful lot of the system yourself anyway. You'll love how it actually works... as far as I remember any code a third party uses has to drive the MSFS built-in autopilot rather than taking direct control of the aircraft. I could be misremembering that but I'm pretty sure that's how it works.
An autopilot is an autopilot, a FMS is a FMS, they all do the same job inn the end with differeint front-ends, so reusing CRJ code for an Airbus or anything else is quite possible.
-
On 9/17/2022 at 4:36 PM, Tomcat_22 said:
The CRJ series has been been out for months and sadly there has been no talk of a fix, I believe the best I read was ''we'll look into it'' then nothing again.
At this point I'm hoping a modder or modding group and sort this out as it seems Aerosoft isn't capable of fixing it, whatever that may cause it.
It's a pity tho, I love flying the CRJ but it just sucks that the LNAV is so bad usually, when it works it's great tho but that's seldom
This isn't the sort of thing a modder can fix unless they somehow decompile a bunch of code libs ( and if they did, they wouldn't dare say anything because that's well outside even generous T&C ). If Aerrosoft are reusing some code for a future project like the Airbus then it needs fixing anyway.
-
2
-
-
13 minutes ago, Tomcat_22 said:
Not easy to reproduce... it's been happen on almost every flight since the CRJ was released, fly into LSZH coming from the north in a north to south wind (so you have to fly past the airport, do a ''U-Turn'' to land on either RWY32 or 34. You'll see it struggle with VNAV
This is about LNAV though, and even for me who's had problems since the start ( a rig upgrade ago ) LNAV issues are inconsistent.
-
4 hours ago, acbent123 said:
Best of luck in your searching. I look forward to the eventual fix as I've been experiencing a lot of the hunting lately. Thanks.
Yes, this seems awkward & may require a whole chunk of code rewritten just to find it, let alone fix it...
It does feel like it could be a gain issue, or possibly a clamping issue in a controller somewhere, although not sure quite how that relates to a 200nm leg causing oscillations without knowing the actual algorithms ( if the a/c is generating a steerpoint, then maybe there's a lot of noise in the generation when there's a long distance ). If that's the case the mystery is why, in a system which is *supposedly* accurately timesliced, some people have it worse than others.
-
1
-
-
38 minutes ago, acbent123 said:
I guess I should have clarified that I don't aim the heading bug directly at the waypoint, but corrected for wind component. Wind shifts and increase or decrease in speed changes obviously affect heading and adjustments are needed. I am inflight now CYFC to CYYZ and the crj is really acting up. I took 10 - 12 oscillations back and forth to finally acquire the track for my first waypoint - worse than it's been in a long time actually. Who knows, maybe this update will help things.
The first WP has always been problematic, that *is* one I've usually flown in HDG. Not aquiring track halfway through the flight is really uncommon though.
-
28 minutes ago, acbent123 said:
I've used heading mode a fair bit too. It's fine if you don't have a strong crosswind component, but it you do, you're soon far off course, or you're having to make constant course corrections. One time, with a +/- 100 kt, nearly perpendicular crosswind, I ended up almost wrapping around a waypoint in an attempt to cross it.
Which is why you don't stick the heading bug on the waypoint, but still fly using your instrument of choice to stay on track. The wind on my run varied from 75 to 150kts with varying crosswind componnts, so it certainly wasn't a case of just sticking it in HDG & leaving it. The CDI bar is pretty handy.
I don't generally have to resort to HDG, this is the first time in a long time, and I can't remember the last time I had to use HDG for so long.
-
OK, there's one that triggered constant overshoot no matter what I did ( fiddled with the stick profile a fair bit too ) - EGPEDCT STN DCT ODPEX DCT ORTAV DCT ASRUN ASRU1H BIKF ( simbrief handed me that one). ORTAV to ASRUN is a very long way, and the -900 just would not settle at all, I stuck it in HDG in the end & tried at various points to put it back in NAV & it only settled down in the last 30 miles or so.
-
1
-
-
On 6/6/2022 at 6:47 PM, Mathijs Kok said:
Part of this is addressed in the update that is being tested. Keep in mind this a 2010 aircraft and not a 2022 Airbus though.
The FBW FMS seems to synthesize intercept waypoints, I don't know what the more default based ones do. For some reason flights into my local airport have been using the ILS approach today instead of RNP ( everything on the ILS app goes over me as opposed to only certain flights on the RNP app, so the noise made me look! ) - the approach is very very similar to that last one I used, and some of the real aircraft appear to cut the corner too ( mostly ATRs ). I think it's fine as long as there's no overshoot - that's not really the time to fly out the other side of the lobes - except of course if the real thing does something else.
-
1
-
-
ILS/NDB procedural approaches seem to be kryptonite for MSFS autopilots at the moment - I've been flying the 146 a bit & noticed some wierdness, so I tried it into EGJJ ILS26 ( which tbh is almost identical to EGTE's ILS26, and quite a few other approaches in the UK at least ). Here's the LNAV part of the plate from current Navigraph AIRAC -
Flying these back to back, same wind on runway heading more or less
* C310 with the PMS50 750 did probably the worst, although at least it stayed in airspace - left alone it did a circle & a half after initial fix, I'm not really sure why other than that it didn't really turn tightly enough & came into the intercept from slightly north. Was about 5 deg off course though, so no excuse.
* 146, and the WT CJ4 - which share the same FMS - both overfly the initial fix, get halfway to the first waypoint and then turn right to go to the second one. Tried without & with the 5% roll nullzone we worked on here.
* CRJ - flew it well, if inelegantly ( with the 5% nullzone ). Turned left from second waypoint until it was pointing at the third rather than flying the corner, which isn't ideal because it can lead to overshooting the ILS lobes but worked out fine in this case. I know it only has two bank angles, but would the real one stop turning halfway round until a second turn would put it on a proper intercept? it did happily fly the LNAV side of a RNP AR approach with curves all the way ( see other thread ), but that one had waypoints all the way down it.
* WT3000 equipped Longitude ( with the little tweak mod ) tracked the corner perfectly.
* PMS50 530 equipped BN-2 flew the corner properly - that's the only one I've tried with that AP & it's got some pretty vicious PID settings. Would have preferred a jet with one so I could fly the same speeds, but I don't appear to have one.
Edit: as expected, FBW 320 - about which I know next to nothing & basically made educated guesses about - absolutely nailed the approach. That one's source code is available too, isn't it?
What is different about the AP in the WT CJ4 to whatever AP the WT3000 and the PMS50 530 uses? source for both WT options is available - I haven't studied it but I bet someone here has. May be clues in there for errant LNAV behaviour? perhaps it's actually an asobo API issue?
-
2
-
-
1 hour ago, acbent123 said:
I have noticed lately that there is about a 14-16 degree wind direction variation between FS (wind speed and direction displayed in LittleNavMap) and the CRJ. For example, in flight now, I am seeing 238/51 displayed on the CRJ screen and it indicates 254/51. The windspeed is always the same, but the direction always varies by the 14-16 degrees. I have no idea if this has anything to do with the hunting behaviour, but thought I'd mention it. Thanks.
What's the magnetic deviation where you're flying, out of interest?
-
17 hours ago, Crabby said:
Well, don't go down a rabbit hole based on my and my son's experience. Like I said, I gave up trying to understand computers a long time ago. I have no clue why some never see this issue, some see it all the time and some see it intermittently. It's a riddle wrapped in an enigma.
No not poking anything deeply, just gathering anecdotal data. I seem to remember the WT CJ4 also at least had a period of similar LNAV problems - that sort of thing makes me consider if they were using similar concepts architecturally which should work in theory, but because something common isn't behaving as it should...
But mostly gathering anecdotal evidence lets you rule things out.
-
1 hour ago, Crabby said:
Never had VR and my son never has either. I have trackir now. The only differences are controllers. He has my old Thrust Master T16000 throttle and joystick and I have the Honeycomb Alpha and Bravo. Both of us have MFG Crosswind pedals. He has the stock keyboard and mouse and I have a logitech wirless keyboard and mouse. The only other difference would be he tends to buy from the msfs store and I avoid it like the plague (don't think this is the issue though), I had my MSFS install in a custom directory until the last beta. When I came off the beta the sim did a full re-install and I did not notice and my community folder changed to the default. I now run my community folder with symbolic links from where I install things to my old community folder, then I link that community folder symbolically to the default install directory after deleting the "real" community folder. I did that so I did not have to make several changes to the folder structure I use to install add ons.
I really doubt install location is going to make a difference - symbolic links ( ala addon linker ) are just a sign saying "things are over there now", they don't change any files - and everything has to be already loaded before the aircraft can operate anyway. Did he get the CRJ off the store too? I guess I should re-ask that as a general question, is anyone who got the CRJ direct from Aerosoft or some other third party that doesn't use the store got the LNAV issue?
I might try removing my TM throttle & the TM software - I guess a straw poll of users & their controllers would not be a waste of time too.
-
1 hour ago, Crabby said:
I gave up trying to figure out PC's a long time ago, just like women. I have the same gaming rig as my son. Exact same. Ordered from the same place on the same date and delivered the same day. We fired them up at the same time. We are both running the same updates. He get's CTD's with MSFS and I have had one that was traced to a bad livery. We have mainly the same mods. I have a few more payware airports because, well I can afford them better than he at this time. Other than that who the hell knows. Maybe his memory sticks are not as solid as mine but they tested out the same. It is nuts. Also, he has reported to me that he sees the LNAV issue. I have never seen it on his system so if he does have it is intermittent. Funny thing is that his rig runs other games like shooters about the same as mine.
I believe that it is the quantum uncertainty principle leaking into our macro world.
That's interesting - do you have the same controllers plugged in? you said you had VR at some point, does he also?
-
2 hours ago, Crabby said:
Computers and code are a weird combination. This whole issue could come down to:
1. an error in code that only manifests itself on certain local system combinations
2. an error in local systems that many folks have without knowing they have it
3. some common third-party mod that causes this
Unless it's navigraph data causing it ( that'd be something freaky ) I can rule out 3, because I have the same issue with or without any mods.
Lets look at what's common & what changes.
* Did this problem not come up in different sims? if so, we can *probably* rule out it being specific to MSFS's api, at least.
* What's different between systems? well, everyone has different drivers & so forth, but does the CRJ code interface with any of them directly? I can't see why it would, it's interface is mostly MSFS because that's what it's controlling. People's systems have different performance and possibly different responses to anything relating to time slicing - and timing is absolutely critical for controllers ( I mean software controllers like integrators, not game controls ). I have coded some basic controllers in the past with unreliable time interval data & they can go really bizarrely haywire. If the code is relying on system time & the parent sim itself is not behaving properly with respect to the passage of time ( have you ever been in a group flight with someone with fps issues? they will appear to slow down so much you risk stalling trying to stay in formation, but their sim is telling them all is fine - I do not trust this thing much when it comes to timing ) then things are going to get wobbly too.
* There could be other ultra bizarre corruption errors caused by something as daft as the host PC's localisation. Or maybe a specific version of VC++ redistributables combined with something *that* interacts with, or something horrible to trace. I suspect the only way to get enough data is to give people who've got an issue a logging version of the aircraft's control library & then have them do specific tests, and hope...
Good luck team!
-
1
-
-
10 hours ago, acbent123 said:
Haha, that was my thread. It seemed to make a difference for me too, but since putting the dead zone in, I still get some deviations.
Oh, haha, so it was
worth publicising for others still I guess. I haven't flown the CRJ much recently but I get roughly the same results every time, especially approaches.
-
1
-
-
On 4/6/2022 at 8:40 PM, acbent123 said:
My flight today was riddled with LNAV issues. The plane either undershot or overshot most of the waypoints and rocked back and forth to the point where I had to diesngage LNAV and go on heading mode and manually get my plane back on track. Can we expect an announcement soon about when we might see a fix? Thanks.
As something of a workaround, can I suggest trying a 5% deadzone in your aileron axis? as detailed in this thread:
it did make *some* difference.
-
1
-
Soooo......any news?
in General Discussion
Posted
Discord is generally a terrible place to post bugs unless there's a specific ticketing system set up because generally issues get lost in noise, but even then most people who do that just use it as an input mechanism for a proper ticket backend. I'd hope there's at least an internal bug tracker.