[time-nuts] TV sync
jpawlan at pawlan.com
Sun Oct 7 12:03:45 EDT 2007
> Hi Tom,
> a related very interesting experiment is to take the Horizontal Sync signal
> from an old Tube-type TV, and feed it to a frequency counter with GPSDO time
> base (preferrably).
> Got to be careful about the high voltages inside the TV though!
> BTW: it's not 15374KHz, there are some significant decimal digits (I don't
> recall the formula to calculate the frequency exactly, but it was the total
> number of frame lines multiplied by 59.94Hz or so).
> Then you can test how accurate the broadcasters' 27MHz reference clock is.
> One would be surprised how inaccurate(!) some broadcasters are, and how much
> drift some Satellite providers have.
> A lot of the stability is dependent on the receiver of course, and how well
> the receiver's VCXO locks to the Broadcaster.
All larger TV stations use Cs standards. What you don't know is that the FCC
assigns SLIGHTLY different scan frequencies to each station on the same
channel in a close area so when you are in a fringe area between two stations on
the same channel, you will intentionally see both pictures superimposed with a
rolling of the scan lines. Otherwise they would be locked and you would see
only black and white bars.
Satellite broadcasts are completely different and I do not know the FCC
specification for those.
More information about the time-nuts