GPSrChive Discussion Forum

LWa wrote: Mon Sep 25, 2023 12:53 pm Richard's RINEX data sample of Garmin GPSMAP 67 (software version 6.60) looks quite promising.

I just wondered... it is possible to use RINEX data to get more accurate co-ordinates of where the GPS was placed? Either through better processing of the signal, or some kind of time averaged position? I was saving the rock as a waypoint and then accidentally saved it twice and noticed that it gave slightly different coordinates (I guess GPS inaccuracy). Not that it matters, I was just curious.

LWa wrote: Mon Sep 25, 2023 12:53 pm Carrier-phase ambiguity fixing could not be demonstrated with this data set. Maybe the data set was just too short or other problems exist.

I could put it on my window sill and leave it running all night if you like

Richard1982 wrote: Tue Sep 26, 2023 9:10 amI just wondered... it is possible to use RINEX data to get more accurate co-ordinates ...

The RINEX files contain the original observations (code and carrier-phase measurements). Combining them with additional data can give much more accurate coordinates. Such data are either similar observations from reference stations, in your area freely available from the Ordnance Survey, or high accurate orbits and satellite clock corrections.

Garmins GPSMAPs produce coordinate results with an accuracy level of a few meters, under ideal conditions. Time averaging can improve this, but you would still be at the 1 m level. With the GPSMAP 66sr it was demonstrated that centimeter level can be achieved in post-processing based on its RINEX data and additional data. The question right now is whether the observations of GPSMAP 67 also have all the characteristics required for centimeter accuracy.

Richard1982 wrote: Tue Sep 26, 2023 9:10 amI could put it on my window sill and leave it running all night if you like

To obtain good quality observations, the GPSMAP antenna must be pointed strictly upwards. There must also be no reflective surfaces above the antenna horizon (no walls, windows etc). So a set-up on a window sill is straightforward, but not the best choice for good quality observations.

RINEX on a GPSMAP 67 sucks. Based on comments in this thread, it appears that Garmin doesn't care.

I have seven consumer-grade devices capable of outputting RINEX data: Four Google Pixel phones (two 6s, a 7, and an 8), a GPSMAP 66sr, a Montana 700i, and a GPSMAP 67. I use open-source program RTKLIB for post processing. With an open sky, the Pixel phones and 66sr usually put 50% of the observations falling within 60mm (2.4") of the average. The Montana is not as good as 50% of the observations usually fall within 110mm (4.5") of the median. OTOH, the GPSMAP 67 sucks. I'm lucky if the readings fall within two meters of the average. The plot of the 67's GPX data is better than the RINEX plot.

Pixel phones are my go-to device for RINEX output.

Any news about Rinex after update to 8.80?

Still terrible with 8.80

LWa wrote: Wed Sep 27, 2023 11:02 am With the GPSMAP 66sr it was demonstrated that centimeter level can be achieved in post-processing based on its RINEX data and additional data. The question right now is whether the observations of GPSMAP 67 also have all the characteristics required for centimeter accuracy.

To temper expectations, you are not realistically going to achieve cm with a Garmin handheld.

The 66sr test was an extreme example conducted by knowledgeable experts. For example, it required a special rotational mount and 16 hours to first determine the antenna PCO & PCV's, and setting up an immediately adjacent base to eliminate atmospheric effects, and the results only started approaching cm after 10 hours observation.

With some knowledge and experience you may achieve decimeter on static points, otherwise a reasonable expectancy is sub-meter.