Small constant LOC deviation during ILS approach

[ghiom 2]

I have searched the forum to find an answer to my problem but could not 100% get it.Here is my problem:Yesterday, I have started flying my PMDG 737-700 again after a long break spent to enjoy some other add-ons (not to mention).I landed three times, each time I shot an ILS approach until 1000' where I disconnect the A/P.The three times, I noticed that after LOC capture, the LOC line on the PFD was not in the middle (although the AP fore sure to track the LOC) and I was slightly right of the runway, by some 1-2 notches. There was no problem with the Glideslope. Obviously, I had the CRS matching the FMC data and both NAVs tuned to ILS. On the three occasions, the winds were light (2-5 knots).I have the latest AIRAC and the latest Sid/stars by Dan Downs.When I used to fly the PMDG some time ago, I never had this problem.There is one thing I have done while flying the other aircraft was to update the magdec.bgl file with a revison found in the AVSIM library (magdec05.zip by Herve Sors).Could this be the culprit?Is the PMDG 737 AP impacted by the informations in this file?Has someone else experienced the same issue?hank you for your support.Ghiom Viguie

[downscc 1,327]

All navdata is based on magnetic bearings, including airac and sidstar formats.There are two aspects to your question; first, FS9 has one bgl file as you mentioned for global magnetic variation. With an update, you are now flying an accurate magnetic course but the installed scenery in FS9 will probably be in error because each airport and navaid bgl has it's own variation value. These values shift at varying rates over time.Therefore, the course error you are seeing between actual track and extended center line is problably still due to the same source. I have observed up to 3 deg differences in LOC course real world verses FS9, and a second source of error is the course provided for the LOC in the airac wpNavAPT data file based on published data that compounds the FS9 error.Thank goodness the only real reason to need localizer course information is to adjust the HSI-type display such that the course is vertical. The localizer has only one course, unlike VORs, and it always leads to the localizer antenna at the far end of the runway (except for a few exceptions like LDAs or one of the JFK ILSs whose antennas are aside the runway).Once you are coupled to the LOC/GS, the simulation is following the ILS provided by FS9 with it's limitations.Footnote: SIDSTARs do not define an ILS course, they may define a fix that is supposed to be on the course and those may be slightly offset to the FS9 course due to the different errors I mentioned above.

[ghiom 2]

Thanks for your reply.Does it mean that the new mabdec.bgl was the cause?I am not sure I understood fully your point.

[ghiom 2]

Just to had some further explanation.What I experienced is an offset LOC track, throughout the approach. The needle stays a few notch from the alignment.Very strange.Thank you.

[ghiom 2]

One more thing to add:When I am on the rwy threshold with the ILS tuned, the needle is perfectly aligned.It looks to me like the A/P is juts not "strong" enough to follow the LOC.But what is very strange is that this is all very new. So far, 737NG was the the best add-on for LOC tracking, even in strong crosswind.Support very much appreciated.Thank youGhiom

[Guest airx3000]

Got the same problem.The course indicated on the PFD always differs from the one indicated by the FMC . Also the FMC is the real course.I didn t have this problem before and i didn't use the .bgl you talked about.

[downscc 1,327]

"the needle stays a few notch from the alignment"Maybe I answered the wrong question.. too much info on mag var.The CDI (course deviation indicator), as I understand it, is centered when sitting on the end of the runway on the centerline but during approach it is off course.. when does the needle change from a few notches to center? How much is a notch, do you mean a needle width or two or as far as one of the four bubbles that are left and right of center on the approach HSI display? Is the flight director in agreement with the autopilot? I assume you are coupled to the LOC/GS and these are annuciated on the flight status display above the attitude indicator.Note that the ILS information presented on the attitude display is derived from FS9 and the ILS information on the APPR INIT page in the FMC is from wpNavAPT airac data. These two will often disagree, so what we end up doing is selecting a course that satisfies the FMC so we don't get the course mismatch error, but the simulation is coupled to the FS9 localizer. Maybe this is the source of the CDI error -- which would surprise me because I always assumed the CDI would center when on LOC course regardless of what the MCP course selection was.Also, please include the airport and approach. I may try to duplicate when I get home this afternoon.

[ghiom 2]

Thank you for all your support.Made a few tests this week end.With updated magdec.bgl, 3 times off-set on 3 differents airports.WithOUT updated magdec.bgl (eg with FS9 original's), 3 times on the spot on the same 3 different airports.It is very clear to me where is the problem coming from.However, I don't know how to expain it.Thank youGhiom

[Guest mcalpinep]

Got the same problem ,though in my case the course is offset by 2 to 3 deg on just about everything (VOR,TO rwy heading etc,747-400 btw) - most noticeable as mentioned above during ILS approach - this happens on multiple flights to every airport (Glasgow INT,Heathrow,Prestwick,Charles De Gaulle). I haven't updated the magdec.bgl file so still using FS9 original , I have updated to the latest airac cycle 701 and I have to admit that I did not have this issue using the originally installed PMDG airac. Silver lining is that I now have to fly approach with manual adjusts after initially capturing LOC - it is helping my pilot skills.

[downscc 1,327]

FS9 bgl files define localizer magnetic course based on a magnetic variation value contain within the scenery file. Separately, the aircraft's magnetic heading is derived from the global magnetic file. I think when you update the global variation, the magnetic course being flow is now different from the localizer magnetic course. Adding to the mismatch, the airac data defines the localizer course based on current published approach data that probably doesn't match either the old FS9 blg scenery or the global FS9 magnetic info. This is an example of MS implementation that could have been done better. Anyway...AFCAD can change the mag variation and localizer course data inside FS9.

[hervesors 1]

Hi all,In FS9 and FSX, magnetic variations as defined in airport BGLs are not used by the SIM kernel for anything else than calculating preferential Rwys for AI traffic and apparently nothing else..This has been discussed a few times in some scenery design forums..Try changing this value to an unrealistic MV with AFCAD and have a look..it will not hurt!So,AFAIK, the magdec.bgl is the only source used by FS for calculating magnetic variation..it will simplify ;-)On a scenery design point of view, the only condition that is necessary for having a correctly centered localizer (on the instrument) is that the TRUE localizer heading equals the TRUE physical runway heading and that the localizer antenna is properly positioned on the same true axe beyond the opposite threshold. Magnetic variation doesn't matter hereAn updated magnetic variation value, will just give here a magnetic heading that will fit more closely the actual published ILS/Rwy heading and the course arrow indicatorHowever, I agree possible problems may arise from:- differences between FS and PMDG wpNavAPT.txt runway coordinates- differences between FS and PMDG wpNavAPT.txt ILS magnetic headings- magnetic variations used in reference charts (some have not been updated since 1995)Hopefully, most problems usually have an explanation ; I will be glad to check for a possible cause and try to provide a explanation if I have the following piece of information:- The airport and ILS used- The loaded PMDG AIRAC number- and optionally a screen shot, showing the problemUntil then, I'm afraid I cannot contribute towards an understanding of the problem. By now, the provided file gave very good fit for most users using charts with 2005 magvars (including PMDG users)Herv

[downscc 1,327]

Thanks, Herve: I'm always learning. I've hand coded airport bgls before and didn't know that about the mag variation variable.Something I've always wondered about are the navaid bgl records for VORs... since these used to be primary navigation when I was hand coding (I'm all for progress). Are the VOR var values used by the SIM kernel or is magdec.bgl still the dominate player. The FAA set's mag variation in the VOR electronics to match that at the time of commissioning, and doesn't change it annually. I'm not sure how often it's done, but as long as airways continue to use radials we really don't care in the cockpit about the error. All we need to know is that a specified radial is the airway or fix bearing. However, flight simming relies on our being able to simulate the VOR.. even it it's variation is established a couple of degrees from the actual.

[ronzie 21]

I'm going to butt in here with a recollection regarding a statement that it can depend on the state of the user aircraft whether a modified AFCAD file ILS is used or the data in the APxxx.bgl. It seems that if you are on the ground one is used but above ground on approach the other one is used.This was part of a discussion regarding why LOCs can be added to an airport in AFCAD but AI will not follow them but user aircraft will as I recall, and that additions in AFCAD itself are not sufficient.If this is true, then you can be off-course on approach but on the ground you will be getting different data when lined up at the threshold.It may have been at the PAI discussions. Include Jim Vile (jvile) and Reggie Fields (rfields) in your search in the general FS9 forum here with the subject ILS and you may find something similar.

[Guest mcalpinep]

>>If this is true, then you can be off-course on approach but on>the ground you will be getting different data when lined up at>the threshold.>Interesting point Ron, what I have noticed when this problem occurs is being 2 deg off course capturing the LOC but when I hit the runway the aircraft re-aligns itself during the rollout.

[Guest corrie]

Ok alot has been written here , the problem I had is when I installed FSX on the same PC as FS2004 it corrupted my FMC display. I would come in for a landing and the runway on the FMC will be far of to the left.You will have to uninstall PMDG and reinstall it.RegardsJohan