What exactly does SV C/No represent in table/statistic/chart views?

0 votes
I have seen the recent post on a similar topic, but am still unclear about this. Am I right in assuming this is a value locally calculated in u-centre from the individual satellite C/No ratios? If so, then what satellites are included? For example is it all received sats regardless of status, or just the ones used in the NAV solution? If it's all satellites then I would expect the figure to jump about quite a lot if a very weak satellite comes in and out of lock and hence be rather unstable. I am trying to fix a poor performance issue with a chip antenna and possibly locally generated noise, and am looking for a convenient single parameter to observe as I make alterations.

Thanks, Alan
by buzz53 asked Aug 8
Refer to several other forum items that describe how to assess C/N0 value.
Presuming you are working with a u-blox product, you should refer to the product's Hardware Integration Manual and Application Notes for guidance.
0 votes
U-center does not modify the individual satellite C/N0 values.  Those values are reported in GSV, SAT, SVINFO messages.
The "avg" values are calculated; min/max are just stored.  The usual computation of these values is based only those satellites used in the nav computation except for the cells in sky view.
The sky view pattern is best for assessing antenna sensitivity pattern if the UUT is mounted in open sky, no obscuration with PCB orientation(with respect to North and horizon)  known and fixed for 24 hours.
Better indicators are in MON HW: AGC, noise, CW jamming and any changes when you enable/disable nearby electronics.
At best, chip antennas are are challenge due to inherent loss from linear polarization and omnidirectional pattern compared to upward view of patch.

The only not-easily-detectable RFI is pulsed jamming--- either CW or broadband...that needs a fast spectrum analyzer. Pulsed jamming may be so short that it doesn't affect C/N0 or AGC or other metrics, but it destroys the data bits in the 50 bps GPS datastream, forcing the receiver to await another 30 second cycle for the damaged data block to reappear.

 Ground plane size and absence of ANY noisy currents through the ground plane are essential to chip antenna success.
by grampy answered Aug 8
by grampy edited Aug 9
0 votes
The Carrier to Noise is a figure of merit computed by the receiver based on results from the correlator bins. The best place to look at this would be UBX-NAV-SAT (for M8 series) or UBX-NAV-SVIN. There are debug messages that can provide fractional measurements rather than the integers you're likely to see out the standard messages.

Sorry don't spend a lot of time with the uCenter table/stats views, tending to process serial streams on the fly, or post-processing .UBX files and graphing things of interest from there.

So the tables don't wing around all over the place I'd expect a column per SV, say 32 GPS, 24 GLONASS, etc and that things come and go as they become visible and cross the sky.

Seem to vaguely recall a single channel test mode, where the input could be driven by a simulator, or indirectly to the antenna via a test chamber.

Other people I've worked with on RF stuff would connect the antenna feed into a Network Analyzer and use that to figure the parameters, and tune the antenna elements. And Taoglas used to offer such services.
by clive1 answered Aug 8
0 votes
I greatly appreciate your advice. Tomorrow I will try monitoring the parameters you suggest. I can see that they are perhaps better for real-time monitoring of noise issues while I make alterations. Once I have minimised the noise, I think I still need some way to check the end performance. I have found that checking lock time and #sats, although the ultimate test, is very time consuming and also very variable from one test to the next. With no change to my test setup, I find astonishing variations with time of day especially. I presume this is due to the changing constellation but was still surprised by the variation (quiet rural location, Europe, 52deg North). I was hoping that the averaged C/No ratio would be a quick indicator of where I stood, and this would be relatively easy to monitor over time.

by buzz53 answered Aug 8
Worked with some GPS guys based out of Northampton many years back.

The constellation is going to rotate over you twice a day, measurements for position accuracy tend to be over 24 hour periods. On occasion done "cauliflower" plots of azimuth vs elevation vs signal strength to understand the masking/performance of an antenna/location.

Out in the sticks, with an antenna breaking the roof line, you should have a pretty good view of the constellation that's well above the 24 satellite target. If you can't see 9-10 at any given time I'd be surprised. An azimuth vs elevation plot should look like a pac-man/pie with a large void in the northern quadrant.

You should still be able to find GPS Planning software that will pull almanacs and show expected coverage at a specific location/time.
0 votes

Adding an answer so I can attach image (41.5 N)

by clive1 answered Aug 8
0 votes

Sky view with C/N0 per sector

Sky view with average C/N0 per Az-El sector

by grampy answered Aug 9
website banner