[time-nuts] GPS ANTENNA

Bob kb8tq kb8tq at n1k.org
Mon Apr 2 10:01:40 EDT 2018


Hi

The problem with “to much gain” is that it is very hard to tell when you have to much gain. 
The system quite happily chugs along reporting fine S/N ratios. Cut the gain by 20 db and
you get the same C/N ratios. The reason is pretty simple - the noise figure is set by the 
front end of the preamp. As long as you have “enough” gain it will dominate. Is “enough”
10 db or 100 db? Either way it’s enough to make that equation work.

The issue comes in mainly because these modules have a limited AGC range (or dynamic
range). For whatever reason, the AGC situation is not reported upstream. If the module has
run out of AGC and is about to loose it …. you simply have no way to tell. 

It’s not just TimeNuts that have trouble with this stuff. Very large / smart / big budget / outfits
run into this stuff as well. The same issues of poorly documented system requirements / 
non-standard system requirements are a major hassle for them. I’ve seen them spend 
major effort because of this stuff. 

So how to work it out? 

Stuff a (dc blocked) variable attenuator in the line. Pick a sat that is reporting something like
35 db S/N. Flip between two attenuator settings every 30 seconds or so. Note any change. 
Do it multiple times with multiple sats. Then move on to your next settings pair. I’d start big,
maybe a 10 db step and go smaller if you can stand the boredom. 

As long as 35 db stays roughly 35 db, you have “enough” gain. Since sat’s move and that 
impacts S/N, there is no exact 0.05 db sort of number. Even looking at numbers over a day
can be tricky. Things are never quite the same today as they were yesterday. 

It does seem strange that you want the minimum gain. It’s not so strange if you dig into things
like land mobile radio design. Your best overload performance is always going to be the 
design with barely enough gain in the front end. 

Lots of fun. 

Bob

> On Apr 2, 2018, at 6:07 AM, ew via time-nuts <time-nuts at febo.com> wrote:
> 
> Question
> Using L/H  what is too much signal 
> Bert Kehren
>  
> In a message dated 4/1/2018 10:54:23 PM Eastern Standard Time, david.partridge at perdrix.co.uk writes:
> 
>  
> just use a bias tee to feed in the antenna volts :)
> 
> -----Original Message-----
> From: time-nuts [mailto:time-nuts-bounces at febo.com] On Behalf Of Azelio Boriani
> Sent: 01 April 2018 23:29
> To: Discussion of precise time and frequency measurement
> Subject: Re: [time-nuts] GPS ANTENNA
> 
> An unusual attenuator with a DC pass.
> 
> On Sun, Apr 1, 2018 at 10:21 PM, David C. Partridge <david.partridge at perdrix.co.uk> wrote:
>> Or use a choke ring survey antenna and an attenuator :)
>> 
>> Dave
>> 
>> -----Original Message-----
>> From: time-nuts [mailto:time-nuts-bounces at febo.com] On Behalf Of Bob 
>> kb8tq
>> Sent: 01 April 2018 14:43
>> To: Discussion of precise time and frequency measurement
>> Subject: Re: [time-nuts] GPS ANTENNA
>> 
>> Hi
>> 
>> Indeed, it is *very* easy to put to much gain in front of a timing GNSS receiver. These beasts are trying to dig out a signal that you can’t even see with a spectrum analyzer.
>> It’s way to far below the noise floor to detect that way. They optimize things pretty tightly to get that done (and to hit a price target ….). Put to much gain in front of them and they get unhappy.
>> 
>> Making this even more crazy, the survey industry standard antenna *does* have a lot of gain. Survey receivers need way more gain in front of them than timing receivers. Put a survey antenna directly on a timing device and trouble will likely be the outcome. Equally, a survey instrument probably will not be happy with a timing receiver.
>> 
>> Why all this nonsense? As far as I can tell, it goes back to how the very early L1 / L2 survey boxes were designed back in the 1980’s and early 1990’s. They made a basic decision to put a lot of gain at the antenna. Motorola came along with their GPS modules later on. They made a *very* different decision about how to distribute the gain. There are very good arguments on both sides for why they did it this way.
>> The bottom line is still - you need to match things up …
>> 
>> Bob
>> 
>>> On Apr 1, 2018, at 2:36 AM, cfo <xnews5 at luna.dyndns.dk> wrote:
>>> 
>>> On Sat, 31 Mar 2018 10:58:19 -0500,
>>> donandarline-Re5JQEeQqe8AvxtiuMwx3w
>>> wrote:
>>> 
>>>> I found a supplier for high quality GPS antennas at a very 
>>>> reasonable price. PCTEL GPSL1-TMG-SPI-40NCB.
>>> 
>>> *** SNIP ***
>>> 
>>> I had one of those on 25m cable, and it worked fine on a Tbolt , 
>>> until i got an active antenna splitter that also had some gain.
>>> Then i had to replace it w. a 26dB version of same type, else the 
>>> "Jackson Lite" was loosing sync.
>>> 
>>> What i mean here, is that you can get too much gain too.
>>> 
>>> Btw: Good price.
>>> 
>>> CFO
>>> Denmark
>>> 
>>> _______________________________________________
>>> 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.
>> 
>> _______________________________________________
>> 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.
>> 
>> _______________________________________________
>> 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.
> _______________________________________________
> 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.
> 
> _______________________________________________
> 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.
> _______________________________________________
> 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 mailing list