Fix height above ellipsoid and undulation for SBF GPS

[makeller1]

As pointed out in issue #23613, the ellipsoid height and undulation are calculated incorrectly. This likely occurred because Ardupilot has an opposite definition of undulation as the Septentrio receiver (and virtually all GPS receivers for that matter). Septentrio defines undulation as the distance from the ellipsoid to the geoid with up being positive (geoid is above ellipsoid). Ardupilot defines down as positive.

This PR reconciles the undulation definition and fixes the height above ellipsoid computed here.

@tridge, @hendjoshsr71 and/or @BluemarkInnovations

Platform
[X] All
[ ] AntennaTracker
[ ] Copter
[ ] Plane
[ ] Rover
[ ] Submarine

[rmackay9]

Hi @makeller1,

Thanks for the PR. Can you change the commit title to start with "AP_GPS:"? Please note that I mean the commit title, not the PR title.

I've added this to the next dev call.

[makeller1]

Hi @makeller1,

Thanks for the PR. Can you change the commit title to start with "AP_GPS:"? Please note that I mean the commit title, not the PR title.

I've added this to the next dev call.

Done.

[BluemarkInnovations]

Thanks for the PR. I made the initial commit, but during the development, a lot of changes were made prior to merging to the master code by multiple developers. I did some research on this and I agree that it looks that the ellipsoid height and undulation are calculated incorrectly. However it boils down how ArduPilot defined undulation. Having said that the current implementation is in any case in conflict with the definition in the AP_GPS code.

So a lot of words to say that I don't have the expertise for this bug/PR, but to me it looks that it is a bug that is fixed with this PR.

[rmackay9]

We discussed this on the dev call (06-Jun-2023) and decided that we needed to see the data from the GPS to be sure this is correct.

@tridge apparently has a HiTec septentrio based GPS that he will test with.

[tridge]

I have a septentrio mosaic here and will test

[tridge]

tested with a comparison of a ublox F9P and a septentrio mosaic on the same antenna, the PR is correct:

[rmackay9]

Great news. I've added this to the Copter-4.4 project. It was added to Plane-4.4 but not Copter-4.4 for some reason.. we should always add to both if we add it to either.

[peterbarker]

Merged, thanks!

[rmackay9]

this is included in Copter-4.4.0-beta3