GPS Failover failed

[KRviator]

'Morning all.

I was out and about yesterday, and for the second flight in a row, my GPS-2020 appears to fail following engine start. That is to say, I have Synthetic Vision available, the clock updates to LMT and XPDR enters GND mode until engine start, shortly afterwards, SynVis fails and the transponder transitions to ALT, which I believe is normal behaviour when it loses GPS data.

The system is a single-screen D-1000 Classic, EMS,  SV-XPNDR-261, SV-GPS-2020 and ARINC-429 module connected to a KLN-90B.

The KLN is configured as GPS 2, with the SV-GPS-2020 setup as POS 1 per the manual.

The problem I have, is with the apparent failure of the GPS-2020, SkyView does not fail over to the KLN-90B,, the map remains static with the "?" centered, terrain data remains red, but the RX Counter for the GPS-2020 keeps increment, suggesting data is being received even though the GPS FIX screen shows no satellites being received. Of note, both the KLN-90 and Ipad running OzRunways had solid GPS lock.

Given SV shows an incrementing counter for the GPS Serial data tells me the connection between the GPS and SV is okay, and as I haven't changed anything settings-wise or elsewhere I'm at a loss as to what the problem may be.

So here we are in flight, it obviously has some kind of GPS data, as it is showing both the wind vector, and current groundspeed.

The SV-GPS-2020 serial screen, apparently in the process of incrementing the RX counter during the photo.

The ARINC-429 setup screen.

 

[Dynon]

That's odd. What does the GPS FIX STAUS page show (note that it will only show info for the SV-GPS-2020, not the ARINC device)?

Also, if you entirely disable the -2020 or the ARINC device in setup, does that change any behavior or give any clues?

 

[KRviator]

Before engine start, it shows the expected green bars and elevation/azimuth of the satellites being tracked. Give it a few minutes and it loses the GPS, and doesn't seem to recover it until the next engine start / power cycle.

With the GPS-2020 serial input selected to NONE, I still have a static map with red terrain warning, but also still have the wind vector and groundspeed readouts available on the PFD, it just doesn't seem to like the Lat/Long input from the KLN to know where it is on the map page...?

With the KLN-90B set to NONE, I lose groundspeed, wind data and the map page fails to a red X, which shows "NO GPS POSITION SOURCE".

I thought it could possibly be related to the GA35 antenna (don't ask me why I have a GA35 with a King GPS, it came with it when I bought it and it works) and the known oscillation failure mode, but that wouldn't explain why the KLN & Ipad retain GPS lock. As a test, I deliberately left the KLN off on the return flight and the SV-GPS-2020 failed again.

Here we are on the ground, before engine start.

With the GPS-2020 selected to NONE, but GPS data still being received from the KLN-90 through the ARINC-429. On another note - any idea why it shows "GARMIN 430 / 420 / 400" in the corner? There is nothing in the setup to suggest that I have connected a Garmin to the ARINC module. And is the loss of AP & Transponder Annunciation in the top bar by design, as I've entered the setup menu?

With the ARINC-429 input set to NONE - note the lack of groundspeed & wind data.

 

[KRviator]

Any updates as to what may be causing this issue, Dynon?

 

[Dynon]

So it sounds like there are two issues here.

First, on the KLN - it sounds like SkyView isn't seeing a complete enough set of ARINC labels for SkyView to solely utilized it as a position source. If you disable the -2020 in serial setup, what does and doesn't work on the map/etc? Are there different ARINC output formats possible on the KLN?

As for the SV-GPS-2020 itself, that's mysterious. What once you've lost the signal, what do you see in terms of movement (numbers changing) on the rx counter, good sentences, and group errors on the serial port setup screen for the -2020?

 

[KRviator]

Sorry for the late reply - got cleaned up by a kid on a dirt bike out front of my house and broke 4 bones in my right arm & leg. Little bastard has stopped me from flying for the last month and probably the next 8 weeks as well...

IIRC, there are no options available to change the speed or anything else of the ARINC interface for the KLN, but I will double-check next time I go to the airport. I can't find anything in the manuals about it.

Page 1-3 of the KLN-90B Installation Manual show the KLN does output present position (Lat/Long) over the ARINC interface using labels 310 & 311, 312 is GS and is received & displayed, and the ARINC-429 Module is supposed to receive these labels as shown in the Dynon ARINC-429 documentation.

From memory, the Rx counter does keep incrementing following the loss of GPS-2020 data, but I can't recall what happens for the good sentences/group errors numbers. Soon as I can get airborne again, I'll check.

This is a relatively new occurance, following installation of the KLN, I did several proving flights with no problems, but in saying that, the SV-2020 was available, so the failover function wasn't tested.

 

[KRviator]

So it sounds like there are two issues here.

First, on the KLN - it sounds like SkyView isn't seeing a complete enough set of ARINC labels for SkyView to solely utilized it as a position source. If you disable the -2020 in serial setup, what does and doesn't work on the map/etc?  Are there different ARINC output formats possible on the KLN?

As for the SV-GPS-2020 itself, that's mysterious. What once you've lost the signal, what do you see in terms of movement (numbers changing) on the rx counter, good sentences, and group errors on the serial port setup screen for the -2020?

'Morning folks.

A bit of a follow-up. The fundamental KLN issue was that I had somewhere along the line changed the Input speed to Low vs High. Resetting that re-enabled the KLN functions, but did not provide position data to SkyView's moving map. As best I can tell, there is no way to change what is output by the KLN-90B's ARINC port, not the speed at which the data is sent.

When the GPS goes down, it is still transmitting serial data to SkyView, there's just no satellites. Not even a hint of satellite reception. If the serial data was not incrementing, I would suggest there is a wiring issue somewhere that is causing the GPS-2020 to disappear entirely, but as you can see from the attached photos, the RX counter climbs rapidly, but the "Good Sentences" seems to stall on loss of GPS reception with the "Group Errors" climbing slowly.

Before engine start, no problems there. The attitude is normal, I'm parked sideways on a hill.

27 minutes later, Message annunciates, with No ADSB Out, GPS Lost then POS1 Fail

No satellits in view, but the RX Counter keeps incrementing

The KLN-90B continues to provide navigation data to SkyView while set as GPS-1, so I still have GS, wind and waypoint data, but no Lat/Long seems to be provided to the Map page, even though it knows where I am supposed to be going, and both the KLN sends it and SkyView is supposed to be able to receive the required labels.

From the KLN-90B installation manual Chapter 1.3.1 shows the KLN-90B outputs the following ARINC Labels that SkyView's ARINC-429 module is able to recieve, but it does transmit several more labels that aren't processed by SkyView.
074 DATA RECORD HEADER
075 ACTIVE WAYPOINT FROM/TO DATA
100 SELECTED COURSE #1
113 MESSAGE CHECKSUM
114 DESIRED TRACK (TRUE)
115 WAYPOINT BEARING (TRUE)
116 CROSS TRACK DISTANCE
121 HORIZONTAL COMMAND
147 MAGNETIC VARIATION
251 DISTANCE TO GO
261 GPS DISCRETE WORD
300 STATION MAGNETIC VARIATION, TYPE, AND CLASS
303 MESSAGE LENGTH/TYPE/NUMBER
304 MESSAGE CHARACTERS 1-3
305 MESSAGE CHARACTERS 4-6
306 NAV/WPT/APT LATITUDE
307 NAV/WPT/APT LONGITUDE
310 PRESENT POSITION LATITUDE
311 PRESENT POSITION LONGITUDE
312 GROUND SPEED
313 TRACK ANGLE (TRUE)
326 LATERAL DEVIATION SCALE FACTOR
351 DISTANCE TO DESTINATION (VIA FLIGHT PLAN)
371 SPECIFIC EQUIPMENT IDENT
377 EQUIPMENT HEX ID CODE

Following on from this, SkyView's ARINC-429 module transmits the following information that is able to be recieved by the KLN-90B, albeit split between 3 input pins:
100: Selected Course
203: Pressure Altitude
204: Baro Corrected Alt
210: True Airspeed
211: Total Air Temperature
320: Magnetic Heading

I have previously flight tested a multi-leg plan from the KLN to SkyView as part of the post-installation checkout and it displayed it on the map without a problem and flew it on Autopilot successfully, so am at a loss as to why SkyView can't seem to recognise where it might be when it loses its' internal GPS. 

 

[Dynon]

So I think it's that (I'll need to check with some of our more deeply technical folks) there's actually not a "complete" gps solution in those sentences. SkyView requires a 3d position, which requires altitude, which doesn't look to be in any of those sentences. The GNS/GTN series actually require an additional connection from a serial port to "complete" the data position, for example. So we consider the other data valid, but not the actual position, and so the only valid position source is the -2020.

 

[KRviator]

So I think it's that (I'll need to check with some of our more deeply technical folks) there's actually not a "complete" gps solution in those sentences. SkyView requires a 3d position, which requires altitude, which doesn't look to be in any of those sentences. The GNS/GTN series actually require an additional connection from a serial port to "complete" the data position, for example. So we consider the other data valid, but not the actual position, and so the only valid position source is the -2020.

That would explain the red terrain warnings, but would SkyView not be able to use its' internal altimeter?

With the GPS-2020 failed and the KLN active, the map page remains available, with flightplan data being received from the KLN, but the "POSITION UNKNOWN" warning active on the map.

GPS-2020 Failed, but the ARINC-429/KLN working:

If I then disable the ARINC-429 module, the map Red-X's with NO GPS POSITION SOURCE, so it looks like SkyView is kinda happy with the KLN, but not enough to provide a position. Reviewing the KLN-90B installation manual shows it can output RS232 data, but it still does not include an altitude datablock in the stream. I am guessing as the KLN - and most other C129 GPS' I suspect - do not output altitude data there is no way to provide failover protection for the map short of installing a WAAS GPS that would output a 3D position?

GPS-2020 Failed, and the ARINC-429 module set to NONE:

 

[dynonsupport]

The map needs altitude to do terrain awareness. You can't use barometric altitude because that can be off by hundreds of feet from reality even when set correctly. The terrain database is all in the same datum as GPS data, so that's the only safe way to use it. The same issue drives Synthetic Vision.

We have the map up with "position unknown" because we are getting GPS data, but it's not "good enough" to yet use it. In normal cases this means the GPS hasn't locked on yet, and GPS altitude will be coming shortly. In your case, it will never come, but there's no way to know that from the data stream we get.

As you have identified, the KLN-90B doesn't output altitude anywhere, so there's no way to use this as a position source in SkyView. It can give CDI deviation, and because it gives ground track and ground speed, the GS and winds on the PFD will work. But without altitude, we won't show the map because we believe so strongly in using the map for terrain awareness, and we can't do that without GPS altitude.

 

[KRviator]

No worries, thanks Dynon. 

At least that clears up the mapping issue. I still have the original GPS-250 at home and a spare serial port so would like to install that, but reading the installation manual, as I have the KLN configured as GPS-1, and that doesn't flag as failed, SkyView would never fall back to the GPS-250 as it would be configured as POS-2 (lower "priority" than the 'apparently-working' KLN on GPS1) - correct?

Which in turn leads to the obvious and final question - any idea what is happening to the GPS-2020??