[time-nuts] LORAN-C demise

Magnus Danielson magnus at rubidium.dyndns.org
Sun Nov 29 22:51:57 UTC 2009


David I. Emery wrote:
> On Sun, Nov 29, 2009 at 12:25:06PM -0800, Mark Spencer wrote:
> 
>> The main difference is that receiving lf signals is challenging in
>> many areas built up areas and the doppler shift of hf via sky wave
>> reduces the accuracy considerably, while there are already a large
>> number of exisiting high power transmitters that can be locked to an
>> external time base and have line of site paths to many locations in a
>> typical metropolitan area.   If a sutiable receiver existed this might
>> be a feasible means of distributing accurate frequency info and then
>> with a suitable reciever you could generate a 1 pps signal.  That being
>> said a dedicated uhf or shf transmiter that could send accurate 1 pps
>> signals (as well as providing a very accurate carrier frequency) might
>> be an easier solution.  In any event if there was a market for such a
>> system I believe it would have emerged by now.   
> 
> 	As I suggested earlier, I believe that relatively simple tweaks
> to a broadcast ATSC transmitter/modulator/mux chain to lock both carrier
> frequency, symbol clock, and say PCR clock to a local cesium standard
> with time of day based (initially)  on GPS would probably be quite
> practical and perhaps even little more than using the 10 Mhz (or 27 MHz
> derived from it) from the cesium as clock input for existing plant and
> setting some firmware settings correctly.

There's a standard for ATSC SFN configuration.

> 	OBVIOUSLY as others have pointed out someone has to pay for this
> even though the actual costs might be very small compared to the other
> operating and engineering costs associated with the broadcast
> transmitter plant.  It is hard to think of a more powerful signal for
> time sync in a metro area...

Unless SFN operation is applied, it is fairly unreasnoble to expect that 
it would happend.

> 	As for receivers, existing ATSC tuner/demod chip sets and a FPGA could
> no doubt supply all the usual timing signals (10 MHz, 1 PPS, time of day
> in some  standard format).   One imagines sub microsecond PPS accuracy
> (once propagation skew is measured) is quite possible.

The propagation delay from the transmitter tower and any delay-offsets 
of the tower would add up.

> 	One would clearly need to use a GPS based measurement to establish
> the propagation based skews... 

Certainly.

>> I imagine you could even design a gps timing receiver that could also
>> receive terristerial signals as a backup, but again it does not seem
>> there is a market for this (:
> 
> 	Network effects apply here - if there is no signal to lock to,
> then there is no market for a receiver - if there is no receiver even rather
> low cost changes to TV plant aren't gonna happen or be justified...

Only within SFN regions the receivers would be usefull.

> 	And I suppose the bottom line is that we'd better hope that no
> natural (or perish the thought deliberate man made) event takes out
> enough of GPS to cause GPS based timing to fail.   And system designers
> had better start thinking about very local jamming of GPS timing
> receivers at targeted sites that might cause a vital system (say public
> safety radio) to degrade or fail.
> 

There are several man-made things which may cause loss of signal, 
including firmware errors, bad antenna connections, water in antenna 
(actually happend to me) etc. etc. There are many reasons for the full 
function to fluke out.

Cheers,
Magnus



More information about the time-nuts mailing list