[opendtv] Re: Bolivia TV standard mystery

  • From: "Albert Manfredi" <bert22306@xxxxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Sat, 23 Dec 2006 15:48:34 -0500

Craig Birkmaier wrote:

Dale wrote:
I may not have explained the Santa Barbara markets (121st
ranking) transmitter site distribution clearly. The NBC affiliate
is located in the extreme north end of the market and the
ABC station is 100 miles to the south, with CBS being about
in the middle. Santa Maria is centrally located and generally
can receive all three, but it deteriorates rapidly as you move
north, south, east or west due to intervening terrain.

Sounds like most of the needed infrastucture is already in
place. In this case, it should be possible to use the three
sticks already in place.

I agree with this part. Use the three sticks for transmitter antennas, and at best you can use the same frequency channels on the two most remote sticks, and a different set of frequencies in the centrally located stick.

That would allow those bent on hype to claim "SFN," where in fact all you're doing is reusing frequencies in two towers that are far apart enough not to interefere with each other. At least, not to interfere where it matters to anyone.

You could transmit all three channels (and others) synchronously from
each stick,

No. And the reason is simple: the towers are too far apart to do this reliably.

A true SFN depends on guaranteeing that by the time the signal of one tower is lagging or leading the signals(s) from another tower(s) by too many usec for the given receivers, there will be a dominant signal strong enough to overpower the long lag or long lead signal(s). Let's say, a dominant signal that is 15 dB louder, or more, than any potential interering signals. In areas where the signal of multiple towers are close in strength, *all* have to be within the echo tolerance of the receivers.

This is true for COFDM as for anything else.

If towers that are spaced too far apart for this to be true, as these three are, then you can try to synchronize the towers to achieve that same effect. The problem is, you are designing this system based on some sort of predicted average propagation contour for each tower. You don't have any actual signal source where you can guarantee that a dominant signal will exist. You are hoping that the north tower's signal will have been attenuated so much before it will interfere with the central stick, for example.

But this stuff is very variable. Weather, time of year, and terrain, conspire to screw up your fine-tuned calculations. So on a good day, you might succeed. On a rainy day, you might fail miserably. And worse, even on good days, terrain will make the fine-tuned calcualtions fail in some parts of the market area anyway.

Bottom line: for SFNs, even with COFDM, the safe bet is to locate the towers of the SFN within a cluster in which no two towers are further apart than the GI allows. Check out real-world examples, in Paris and Berlin. In these cases, once you get out to the far field, say starting at 20 miles from the SFN center, the SFN will behave very much like a single stick. Make it low and low powered, and you'll lose the fringe areas completely, just as you would with any small stick. And no, you can't use OCRs effectively at the fringes. Not unless terrain offers signal blocking where it's needed. Which is why, in those real-world examples, they make use of translators to provide signal to outlying areas.


From photos to predictions, The MSN Entertainment Guide to Golden Globes has it all. http://tv.msn.com/tv/globes2007/

You can UNSUBSCRIBE from the OpenDTV list in two ways:

- Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word 
unsubscribe in the subject line.

Other related posts: