[time-nuts] How can I measure GPS Antenna quality?
kb8tq at n1k.org
Mon Nov 21 07:15:08 EST 2016
At spectrum analyzer bandwidths, the GPS signals out of the antenna are
more than 20 db below the noise floor. You can’t see them with an analyzer.
You need to run things into the equivalent of a receiver to turn it into anything
you can see above the noise.
What you will see on an analyzer is the noise out of the preamp. That will at
least tell you that the output stage on the amp is working. It will tell you very
little about the input stage to the amp and very little about the antenna on
the other side of the preamp.
One thing you want to know about a GPS antenna system is the
stability of it’s phase center as things change (like sat angles). If the phase
center moves, your solutions change between satellite readings. Another
thing you want to know is how it rejects multipath. Neither one is easy
to measure in a direct way.
If I put up a handful of antennas on the back porch, I can indeed hook them
up to various receivers and cables. I can take a bunch of data. There will
always be very real questions about location A being the same as location B.
You can swap all the antennas between all the locations and take weeks of
data in each location combination. You have fewer questions, but there are still
A more formal method of testing would be to use a proper antenna test setup.
Those normally are indoor systems that rotate things to test the antenna. Even
there, the systems are only so good. Another proposed solution is to run the antenna
in the real world on a robotic arm and rotate it while in use. Again there are limitations
to the process.
Lots of approaches, lots of questions ….
> On Nov 20, 2016, at 10:22 PM, Mike Cook <michael.cook at sfr.fr> wrote:
>> Le 21 nov. 2016 à 02:57, Tom Van Baak <tvb at LeapSecond.com> a écrit :
>> Hi Hal,
>> That's a very sensible question. I've often wondered the same, but I'm embarrassed to say I have never done a thorough job with it. You know the constellation repeats approximately every 24 hours so you want your X hours to be a multiple of days.
>> Looking at SNR seems obvious and may even be sufficient. Alternatively you could track the deviation of the per-SV timing solutions and draw conclusions from that. I suspect multi-path effects would show up in these residuals more than they would show up with just NSV (number of satellites received) or SNR (signal to noise ratio)
>> But in some respects, the bottom line is not NSV or SNR or multi-path or anything like that. What counts is only how well the 1PPS matches a local high-quality time standard (e.g., Cesium or better).
>> Another issue is that it's possible that the quality of a set of N antenna would sort differently for you than for me: different latitude, different sky-view, different weather. Some time nuts (not me) get lucky with a perfectly clear 360 degree horizon view.
>> I agree that N antennas and N receivers makes the experiment easier, because you might spend as much time validating that a set of receivers are all the same as later comparing various antennas.
> Sensible question but not easy to answer directly without a spectrum analyser directly connected. Not having one of those I went the n receivers/ m antenna method since all I wanted was to get the best subset from what I had. My antenna are of the €5-€25 puck variety and I tested about eight 5V, 5/3V active Noname and Trimble antenna with half a dozen receiver types. I am unlucky in having just a north looking sky view and in built up area which gets me significant multi path at certain times. So I first of all selected a period during the day where all receivers were reporting best SNR and max NSV and most stable 1PPS . I then measured the 8 antenna over the same time interval (IIRC it was 08h-11h) against the 1PPS of a PRS10 rubidium ref. The GPS 1PPS was verified to see if it held the manufacturers stability spec which they mostly did. It was easy to see this against the rubidium signal. So I picked the best 4 and use them to feed my receiver pen via Mini-Circuits distribut
> ers ( There is a small signal loss here but in spec. 3db IIRC ). They have been in place for about 3-4 years at least without issue.
>> With that in mind, consider the slice & dice (chopper) idea -- use a VHF relay switch / mux and round-robin N antenna across N receivers every, say, 10 minutes. That gives you 144 points per receiver / antenna pair per day and avoids the geometry and pre-calibration issues, as well as environment and local reference effects. Rubidium would be sufficient. It's possible the first minute of each segment may be weird (as the receiver switches from lost to lock mode), but you can handle that in your data reduction.
>> ----- Original Message -----
>> From: "Hal Murray" <hmurray at megapathdsl.net>
>> To: <time-nuts at febo.com>
>> Cc: "Hal Murray" <hmurray at megapathdsl.net>
>> Sent: Sunday, November 20, 2016 2:13 PM
>> Subject: [time-nuts] How can I measure GPS Antenna quality?
>>> Is that even a sensible question? Is there a better way to phrase it?
>>> The problem I'm trying to avoid is that the weather and the satellite
>>> geometry change over time so I can't just collect data for X hours, switch to
>>> the other antenna or move the antenna to another location, collect more data,
>>> then compare the two chunks of data.
>>> The best I can think of would be to setup a reference system so I can collect
>>> data from 2 antennas and 2 receivers at the same time. It would probably
>>> require some preliminary work to calibrate the receivers. I think I can do
>>> that by swapping the antenna cables.
>>> If I gave you a pile of data, how would you compute a quality number? Can I
>>> just sum up the S/N slots for each visible/working satellite?
>>> These are my opinions. I hate spam.
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
> "The power of accurate observation is commonly called cynicism by those who have not got it. »
> George Bernard Shaw
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts