So the question becomes is the facet antenna better than the l5 antenna and that’s where you can get down from 22mm with postcard on orgn to 14mm with facet. Or is the formulation of the nmea message not as good. An easy way would be if I had a zed chip to hookup to my l5 antenna. Or I guess maybe I can open my facet and plug my antenna in
I could use a Surveyor or Express, to test F9P and LG290P on same antenna.
Of my several dozens of antennas, I don’t have the one you are using for your PostCard 
I never put much confidence in the “Uncertainty” metrics anyway.
As you said, repeatability means much more, and is what we actually measure/track.
1 Like
I’ve seen every GNSS receiver make the wrong guess on Integers and produce a bad position while still showing RTK fixed with a normal looking RMS.
1 Like
Again, self-reported uncertainty numbers from the receiver itself are an estimation since it does not have enough information (a reference point) to do the exact calculation. It makes no practical sense to assign more significance to it than it is meant to have, it’s also simply metrologically incorrect.
2 Likes
@Bushman_K I really appreciate you helping put my feet back on the ground 
@rftop to that point thats what has made me uneasy about say the Trimble r12i and it showing low RMS and being told that is gospel. I’m sure Trimble has some filtering regimes rtk everywhere may not atleast their marketing would make the case for it. However with commerical units I feel like a lot of surveyors just blindly trust the collector told me so and so I can take 10 second observations and recieve highly accurate results.
To be completely fair, specialized equipment for geodetic measurements might use additional information to estimate error.
For instance, the most generic receiver has no way to differentiate between the drifting solution caused by reception errors and the actual small movements. So the best thing it could do is guess.
A slightly more advanced receiver might have multiple algorithm presets (e.g. um980, um982) that you can choose from to inform it about the actual scenario. For instance, a “base” mode tells it to treat any observed velocity as an error, while the “rover” mode removes this restriction.
A combination of a receiver and a controller/collector might be able to utilize user input as well as additional sensors (magnetometer, accelerometer) for that purpose. Plus, when it’s a self-contained unit with a built-in antenna, it might have proprietary antenna calibration data that allows it to estimate the actual reception conditions (sky obstruction) based on the ideal conditions.
If any manufacturer ever bothered with stuff like that (again, I’m only hypothesizing - I’m not a surveyor but an engineer), it might make their error estimations much closer to reality and more meaningful. Yet, it’s still an estimation.