[time-nuts] TV Signals as a frequency reference

jimlux jimlux at earthlink.net
Fri Mar 30 21:23:18 EDT 2018


On 3/30/18 5:52 PM, Bob kb8tq wrote:
> Hi
> 
> 
>> On Mar 30, 2018, at 6:13 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
>>
>>
>> fgrosz at otiengineering.com said:
>>>   Now that analog TV has gone away, so
>>>   have these signals.
>>
>> What do the local TV stations use for a frequency reference?
> 
> Anything from a crystal oscillator to a Cs standard. It’s very much a “that depends”
> sort of thing. If Crazy Bob is the chief engineer it might be a hydrogen maser ….

And Crazy Bob can convince the owner of the station that it's needed<grin>
> 
>>
>> Are there low cost receivers that also produce a good reference frequency?
> 
> As noted earlier, color burst references were a big deal a long time ago. Depending
> on how they do what they do it might still be a good bet. The big risk is that it could
> be a good bet “most of the time”.
> 



I wonder how stable the underlying timing of ATSC or DVB-T is?  You 
could recover the carrier or bit clock from an over the air signal, 
should you be lucky enough to live where the signal exists.   It's non 
trivial - all modern receivers do it as part of a single cheap 
monolithic chip - but maybe you could find some SDR code to run on a 
PLUTO or other cheap SDR that lets you "see" that level of the signal.

There's no inherent reason why it should be controlled well, at least 
for ATSC - the receivers are designed to tolerate multipath, Doppler, 
and other impairments.

But for simulcasting, the various transmitter carriers need to be 
matched fairly well.


More information about the time-nuts mailing list