[time-nuts] LORAN-C demise

[Lux, Jim (337C)]

On 11/29/09 11:40 AM, "Mark Spencer" <mspencer12345 at yahoo.ca> wrote:

> Any thoughts on how complex a receiver would need to be to produce a 1 pps
> signal that was locked to the carrier frequency it was receiving ? Lot's
> of comercial transmitting equipment is designed to use an external frequency
> standard and if a transmitter at a high altitude site was locked to a cesium
> source it could serve a typical metroplitan area.   Locking an existing
> transmiiter to a cesium standard would not require any special signals or wave
> forms to be transmitted.  To be usefull the receiver would need to produce a
> standard 1 pps output.
> 
> Stanley Reynolds wrote:
>> How about the Volunteer Association of GPS Backup for Timing, VAGBT ?
>> Propose of the group is to provide backup distribution of timing information
>> for GPS users, via armature radio and cesium clocks. To develop many local
>> transmit stations as possible and low cost receivers with both extended
>> holdover and comparison to GPS to measure backup accuracy. Many low power
>> transmitters would be required as the cost of continuous operation would be
>> lower for each station, and the identification of less accurate stations
>> possible if several in each location was avabile.





And how is this is different from time stations like WWV or WWVB?  They're
driven by cesium clocks (or an ensemble of clocks).  The atmospheric
propagation uncertainty means that the received instantaneous frequency
might be off by 1E-7 or so (for HF WWV, at least), but I would imagine that
averaged over a long time, it's quite a bit better (one NIST doc says 1E-9),
but apparently it's tough to do straight averaging. The station clocks at
WWV is is good to something like 1E-13  (adev of 1E-13 at tau of 10,000
seconds, down to about 2E-13 at tau of 1E6 seconds)

WWVB at 60kHz is different.. It's from the same master clock, but the NIST
doc says that received phase is stable to 1E-8 at tau of 2 seconds, down to
1E-9 at a tau of 1000 seconds, with a WWVB disciplined oscillator getting
down to around 1E-12 for averaging over a day (which NIST says is about 1
order of mag worse than a GPS disciplined oscillator)

Of course http://tf.nist.gov/general/pdf/1969.pdf says they use common view
GPS to sync the WWV transmitter to the clocks in Boulder. It doesn't use GPS
for time, just as a source visible to both at the same time)  The same
document claims that 0.1 millisecond absolute time uncertainty should be
achievable with the received signals of WWV/WWVH and perhaps <1 microsecond
for WWVB.

One can propose, say, VHF or UHF signals radiated from a high location which
would potentially have better instantaneous frequency stability (because the
propagation is more stable), but at some point, you're still going to have
to deal with things like SNR and propagation.  Granted a lot of those issues
are "solved" in some sense (e.g. You could set up a GPS pseudolite, like
they do for approach/landing navigation experiments)