[Test-Equipment] GPS locked base ?

[Dave Emery]

On Sun, Dec 28, 2003 at 10:59:04AM +0100, TK5EP (Patrick Egloff) wrote:
> Hi All,
> 
> Is anyone using a GPS locked frequency standard ?
> 
	I am using both a rubidium Datum standard locked to GPS and
several Symmetricom 58540A GPS locked quartz standards.   Both
were purchased on eBay for a couple hundred dollars per...

	The 58540As are small units about the size of a half height
CDROM drive with an internal high stability quartz oscillator and
a 12 channel GPS - they take 24 volt power...  They are wonderful
little boxes and highly recommended.

	As far as I have been able to tell using an ex-Enron GPS telcom
timing box to cross check the 10 mhz coming from the Symmetricom
it is sitting within parts in 10^12 of correct frequency and within
a very few nanoseconds of the correct time of day on its 1 PPS output.

	The Symmetricom has phase coherent 1 PPS and 10 mhz outputs
and slews the 10 mhz quartz standard it contains to track GPS timing
using a narrow bandwidth PLL - because of this it never makes jumps
in either 1 PPS or 10 mhz phase (except when it initially locks to
GPS it slews the 1 PPS) and I have found its outputs are really
stable with essentially no measurable jitter relative to a free
running rubidium oscillator.

	Some of the other GPS gear I have (Datum and Garmin) adjusts
the phase of its 1 PPS to track GPS timing in various relatively
course steps (7 ns or even 42 ns) and such steps occur suddenly
and randomly.   Their 1 PPS outputs can wander around up to maybe
50-100 ns as satellites appear and disappear.



> I would like to have a good reference or at least a precise method to
> use my equipements.

	The Symmtricom 58540A and the very similar  Trimble Thunderbolt
were designed to supply timing to CDMA base stations and are small
conveniant, low hastle solutions with high performance.

> 
> Has anyone used the 1s clock coming out of a GPS receiver as a
> frequency meter gate ?? My idea is to use this 1s rate, eventualy
> divided by 10 to have a 10s rate, and feed it in my frequencymeter to
> have a precise gate rate. Than measuring a 10 MHz TCXO should give
> 10000.0000. If not would mean that the TCXO has moved and i can correct
> it to show the right frequency and have a precise reference. At least at
> the time of measure.

	The best way to use GPS timing is usually to acquire one of
the disciplined standards (like the 58540A) and use its 10 mhz output
as external clock for a counter.   Most lab counters are designed to
be able to use external 10 mhz clock for their timing.

-- 
   Dave Emery N1PRE,  die@dieconsulting.com  DIE Consulting, Weston, Mass 02493