[time-nuts] V standards

[Tom Van Baak]

> Does this mean that by "using carrier phase disciplining techniques"  
> One could get two orders less noise than  is shown 
> on the SRS page#3  graph http://www.thinksrs.com/downloads/PDFs/Catalog/PRS10c.pdf

Warren,

Yes, high-end survey equipment and metrology-level GPS timing
receivers are based on carrier phase rather than code phase.

But you'll rarely see this implemented as a GPSDO (I know of one
and it's obsolete). Instead, for this class of receiver the local lab
frequency standard (e.g., cesium or maser) is used to drive the
receiver's clock and what you get out are a series of internal phase
measurements. These are often batched up daily and then other
non-realtime corrections are applied. What you get in the end is
a precise measure of how good your local standard *was*.

There's no reason in principal that a carrier phase GPSDO would
not work. But I think you quickly run into other limitations that
dominate the net accuracy. Like SV clock errors, orbit errors,
and ionospheric corrections, etc. For fullest accuracy these all
have to be applied after-the-fact (because they are not known
well enough in realtime).

Remember also that the GPS signal, as received by you (no
matter how fancy the receiver), is not a perfect copy of UTC
time as transmitted by the satellites, which in turn is not a perfect
copy of UTC(USNO), which in turn is not a perfect copy of UTC.

/tvb