[time-nuts] GPS Antennas

[Attila Kinali]

On Thu, 29 Aug 2019 12:59:06 -0400
Dan Kemppainen <dan at irtelemetrics.com> wrote:

> For the second summer in a row we were smacked with lightning. So I'm in 
> need of a new GPS antenna. Might as well look for one with multi band 
> units. I recall some talk about the cheap multi-band units on ebay that 
> that are compatible with the f9P modules. Can anyone report back how 
> they worked out?

As a general answer, as this kind of question pops up again and again:

For time-nuts home use, any GPS antenna with a suitable band-filter
and a decently shaped radiation pattern is good enough.

The band-filter is nothing but a (cheap) SAW filter in the antenna,
so that's easy to take care (especially in China). The radiation
pattern is a bit harder, but most antenna architectures are quite
benign. Especially if they are not optimized to be ultra small.

Most people will cite stuff like phase center variations. But honestly,
most time-nuts don't have the means to measure that accurately.
E.g. a bad antenna has a phase center variation of 1-2cm. Let us assume
it's 3cm, to make numbers simple. That would mean a worst case time
offset for a single satellite of 100ps. Ie that would result in a worst
case error (assuming there is no averaging, which isn't true) of 100ps.
Compare this to the (AFAIK) best acheived time-transfer uncertainty
using GNSS common view with integer-PPP is ~180ps. Most GPSDOs we have
wont be even that good and are rather in the 1-10ns offset and TDEV
in the order of 1ns.

				Attila Kinali
-- 
It is upon moral qualities that a society is ultimately founded. All 
the prosperity and technological sophistication in the world is of no 
use without that foundation.
                 -- Miss Matheson, The Diamond Age, Neal Stephenson