GNS530 V2 vertical guidance not working (resolved)

[fppilot 4,355]

23 minutes ago, RXP said:

However it may be in the FltSim version, given the way we override the coupling, it is a dog chasing its tail. I'm reviewing all this now.

Amid all this press the APR button twice if above the glide slope... keep in mind there have been instances of  sudden unexplained nose down plunges, and two other occasions where on final I could not make any altitude related adjustments on the AP.  C210.  Could not switch ALT mode on or off, could not adjust altitude level or VS rate. 

Since GNS 2.4.9 it appears all my flights have been GNS, much due to chasing this.  Is there any similar issue with the GTN release?

Edited March 16, 2019 by fppilot

[Spug 7]

Regarding when to activate the vertical guidance: I wait till I see the glideslope indicators on the HSI. I always make sure to be under the glideslope when the indicators appear that way you "fly into" the glideslope. You're in NAV mode before that, then when the g/s indicators appear, press <APR> and it should maintain the current altitude and start the descent when intercepting the g/s. Works the same way for and ILS approach. I you are above the g/s even though you are on the localizer in NAV mode, pressing the <APR> will disengage the ALT. 

Not sure why after pressing the <APR> the first time, makes the autopilot think you are above the g/s and the second time it works correctly. The entire time you are under the g/s.

At least for now I have a work-around.

[fppilot 4,355]

1 hour ago, RXP said:

the GPS output the necessary signals and these signals drive the FltSim A/P coupling.

By saying FltSim A/P coupling you are saying A/P coupling for vertical guidance, correct?   At what point in the approach is your logic first making the determination of above/below GS and coupling for vertical guidance?  It should be at the FAF and not before, correct?

[RXP 2,338]

@fppilot both GTN and GNS V2 autopilot gauge code is the same so both are equally affected the same way, minor some 'nav' sequencing differences affecting mostly 'when' signals occur. With these signals we feed/override the simulator.

As for the 'RXP logic' there is little if not none: it is solely a Garmin logic determining what to output and when to sequence it. Our code intervenes only to make sure the output signals from the GPS makes sense as input signals to the simulator (for example don't override the F/D if F/D mode is off). Comparing your reports and the code, I may have found a logic flaw though so I'm spending the week-end on this very closely.

@Spug this makes sense to me. I've never really paid attention to the ALT mode disabling before though, because the aircraft was flying level until intercepting in all our tests so far.

Edited March 16, 2019 by RXP

[fppilot 4,355]

1 minute ago, RXP said:

 I may have found a logic flaw though so I'm spending the week-end on this very closely.

Is this only GNS, not GTN?

 

[RXP 2,338]

3 minutes ago, RXP said:

both GTN and GNS V2 autopilot gauge code is the same so both are equally affected the same way, minor some 'nav' sequencing differences affecting mostly 'when' signals occur

If it is ok with the GTN but not with the GNS, maybe some of these signals occurring before or after make a difference. I'm looking very closely to all this to make sure.

Edited March 16, 2019 by RXP

[fppilot 4,355]

15 minutes ago, RXP said:

If it is ok with the GTN but not with the GNS

I have not yet flown with GTN since before 2.5.14.

[RXP 2,338]

@fppilot if you can do the same test flight with the GTN 2.5.14, maybe this would reveal whether there is commonality in the issues henceforth in the logic.

[fppilot 4,355]

2 hours ago, RXP said:

@fppilot if you can do the same test flight with the GTN 2.5.14, maybe this would reveal whether there is commonality in the issues henceforth in the logic.

OK, but probably not until sometime Monday.

[fppilot 4,355]

Yes. GTN 2.5.14 has the same characteristics in regard to coupling vertical guidance  for RNAV/GPS approaches.  I do not have the same C210 set up with a GTN.  I however flew the same approach with two of my GTN equipped aircraft, the Realair Turbine Duke v2 and the Milviz Cessna 310R Redux. 

In the TDv2, descending from 4000 at the IAF to arrive at a target altitude of 2100 before the FAF, and acquiring an LPV indication on the GTN, pressing the AP APR key armed APR mode and VS, but disarmed the AP ALT function.  I continued the approach and there was no vertical guidance.  The AP/GTN never leveled at the 2100 ft target altitude and did not acquire LPV guidance at the FAF.

I flew two additional approaches with the GTN equipped TDv2, beginning the approach just as above in the first approach.  Having learned from Spug's experience, upon acquiring the LPV indication and pressing the AP APR key, when the ALT disarmed I was able with additional AP key presses to get the AP to the right state for following the approach.  APR, ALT, VS and arm indications.  For both of these successful TDv2 approaches the AP key presses were trial and error.   It was not as simple as pressing the APR key twice, and I was not able to map a sequence to follow for future flights. It was try this, try that, until I got the AP indications I was looking for.  For me that is not a satisfactory result.

I then used the same approach with the Milviz C310R Redux model.  The KAP 140 AP in the Redux is a more sophisticated, shall we say, instrument.  I was not successful with either approach in the AP/GTN following vertical guidance. For this Redux KAP 140 AP it for some RNAV/GPS approaches requires establishing on the final approach course with HDG mode, then selecting APR mode after gaining acquisition of GPS LPV guidance.  That adds another layer of complexity to this current GTN and GNS issue that I just could not solve with two approaches.  I then attempted a 310R  approach to my familiar KBMG RNAV/GPS 35 with similar results. 😖 Not happy.

[RXP 2,338]

@fppilot Thank you for the additional reports. They seem to confirm our findings during the week-end.

A quick update: I may have found out what is causing this and I'm doing my best to have a test build today or tomorrow to try. There was definitely a cross-over of code between our XPlane and FltSim version which is impairing each in a certain way. This plus some idiosyncrasies in each platform autopilot implementation which cannot be 'just abstracted out' as we've done in the last 2 or 3 GTN/GNS V2 versions.

[Spug 7]

Looks like v2.4.10 has fixed the coupling problem. Great job. Thank you !!!