P3D V3 FMC memory leak back?

[Skywolf 992]

 

 

I've isolated the leak with KSEA ILS 16L to the following code:



TRK 131 INTERCEPT RADIAL 161 TO

 

Is that from your internal coding? I wonder why VAS leak happens on a very specific code.  Looking forward on hearing your results

[CaptKornDog 282]

Is that from your internal coding? I wonder why VAS leak happens on a very specific code.  Looking forward on hearing your results

No, this is nav data.

[Skywolf 992]

No, this is nav data.

 

Wow - does this mean there are some issues in Navigraph airac data like formatting bugs?  I assume that is the navdata

[downscc 1,327]

Wow - does this mean there are some issues in Navigraph airac data like formatting bugs?  I assume that is the navdata

No, there is nothing wrong with the syntax used by Navigraph.

 

Full names on all posts in the PMDG forums please.

[CaptKornDog 282]

Wow - does this mean there are some issues in Navigraph airac data like formatting bugs? I assume that is the navdata

No, it's how it's somehow processed specifically with the 777 (not on the NGX, at least from what I can tell).

[pcubine 143]

Is that from your internal coding? I wonder why VAS leak happens on a very specific code.  Looking forward on hearing your results

You might want to read these threads so you don't need to ask questions that have already been answered - http://www.avsim.com/topic/455526-777-oom-depleted-vas-my-brand/, http://www.avsim.com/topic/467611-27apr15-ah-that-pesky-memory-leak-and-a-few-other-small-items/, and http://www.avsim.com/topic/468259-07may15-detailed-discussion-of-resolved-777-vas-leak/.

This Forum requires full names on all post - http://www.avsim.com/topic/447430-blunt-warning-sign-your-posts-or/ or the might be deleted by a moderator.

[bjohnsonmn 0]

Knowing the level of detail CKD puts into these tests, I believe he is correct.  I wonder how many other ones similar to this are hiding out there.

 

Regression testing is hard!

 

Best,

Brad

I've isolated the leak with KSEA ILS 16L to the following code:

TRK 131 INTERCEPT RADIAL 161 TO

 

 

 

Without that code in the KSEA file:

 

[PhilH 15]

Xander,

 

I will be sending you a modified KSFO.txt shortly with instructions.

 

Just wanted to say that I was trying to fly LFPG-KSFO overnight while having already programmed 28R ILS in the FMC just after departure. I had a CTD with no warning so I reloaded my save, paused the flight sim I saw my VAS going down even with the sim paused with nothing happening... until I deleted the missed approach instructions for 28R, then the leak stopped.

 

P3D v3.1, PMDG 777-300ER, nav data AIRAC-1513. You probably already knew this but I'm just confirming...

[PinkPony 171]

 

 

I've isolated the leak with KSEA ILS 16L to the following code:
 
TRK 131 INTERCEPT RADIAL 161 TO

Awesome find Kyle! I was able to prevent ILS16 leaking at KSEA by removing the entire missed approach code. Its great to see that you identified a specific string in the missed approach code. 

Would you mind testing "HDG 131 INTERCEPT RADIAL 161 TO".

Have you reported your latest finding to PMDG yet? This may help them narrow it down!

 

Thank you for your continued efforts! 

 

 

Just wanted to say that I was trying to fly LFPG-KSFO overnight while having already programmed 28R ILS in the FMC just after departure. I had a CTD with no warning so I reloaded my save, paused the flight sim I saw my VAS going down even with the sim paused with nothing happening... until I deleted the missed approach instructions for 28R, then the leak stopped.
 
P3D v3.1, PMDG 777-300ER, nav data AIRAC-1513. You probably already knew this but I'm just confirming...

Yes Phil, just as I suspected when I first read your thread regarding the KSFO OOM!

I would like to see if the same line of code which Kyle has identified as being the culprit in KSEA ILS16L is present in KSFO ILS28R.

Specifically, the TRK *** TO INTERCEPT RADIAL *** TO. If so I will send you a updated navdata once again, removing that specific line and you can test that out again.

Hopefully we can identify the specific navdata code responsible.
 

[CaptKornDog 282]

Have you reported your latest finding to PMDG yet? This may help them narrow it down!

 

Yes, my results are similar to those I reported approximately a year ago. I'm working to isolate and compare the other airports, contrasting them with the repaired KLAX.

[CaptKornDog 282]

Data for ILS 28R at KSFO.  OOM a little after 8 hours.  Again, this is FSX SP2.

 

Missed approach code:

RNW 28R HDG 284 UNTIL 420 TRK 251 INTERCEPT RADIAL 281 TO FIX VIKYU FIX VIKYU AT OR ABOVE 3000 HOLD AT FIX VIKYU RIGHT TURN INBOUNDCOURSE 281 LEGTIME 1

And isolated it to:

TRK 251 INTERCEPT RADIAL 281 TO

[CaptKornDog 282]

Switching the KSFO ILS 28R missed approach code from:

 

 

RNW 28R HDG 284 UNTIL 420 TRK 251 INTERCEPT RADIAL 281 TO FIX VIKYU FIX VIKYU AT OR ABOVE 3000 HOLD AT FIX VIKYU RIGHT TURN INBOUNDCOURSE 281 LEGTIME 1

 

to:

 

 

RNW 28R HDG 284 UNTIL 420 HDG 251 INTERCEPT RADIAL 281 TO FIX VIKYU FIX VIKYU AT OR ABOVE 3000 HOLD AT FIX VIKYU RIGHT TURN INBOUNDCOURSE 281 LEGTIME 1

 

(Note the difference from TRK to HDG)

 

resulted in an OOM as well.

 

[PinkPony 171]

 

 

resulted in an OOM as well.

Thank you Kyle, that also confirms my findings in P3D.
 

Looking at the offending KSFO ILS28R code:

TRK 251 INTERCEPT RADIAL 281 TO 

and the offending KSEA ILS16L code: 

TRK 131 INTERCEPT RADIAL 161 TO

I notice for both  approaches the different between the TRK *** and INTERCEPT RADIAL *** TO is a "positive" 30 degrees.

Looking at the problem free KIAD ILS01R (no leak), a approach that I have done dozens of times on 13-14 hours sectors from OMDB:

TRK 081 INTERCEPT RADIAL 051 TO

Here you can see a difference of "negative" 30 degrees. 

I am curious to see if this may be playing a role in causing the leak in the PMDG B777 FMC.

 

 

[CaptKornDog 282]

Tough to say because KLAX ILS 25L, which didn't used to work on my system, but now does after the latest patch, ultimately ends in a right turn to join the radial.

[CaptKornDog 282]

Another data point.  Starting on EDDF runways.