AVS Forum banner
Status
Not open for further replies.
1 - 5 of 5 Posts

·
Registered
Joined
·
304 Posts
Discussion Starter · #1 ·
I think I have a multipath issue, but who knows?


Multipath for analog causes ghosts. You can tell

how much longer one path is than another by the

horizontal separation of the images, right?


Are we talking about (for digital) reflections

from my own roof, a building or hilltop down the

block, or an antenna tower or geographical

feature miles away, or what? Every system must

get multipath signals from things everywhere, it

would just be a question of how strong (normally

too weak to matter) or how delayed (normally so

little or so long as to not matter).


In my case there is a large antenna tower about

3 miles away, pretty much right between me and

the broadcast towers which are about 7 miles

away. Could that be a source of multipath

problems?
 

·
Registered
Joined
·
1,273 Posts
You basically have it right. As for magnitudes, on an analog picture, it takes about 65 microseconds for the electron beam to sweep horizontally across the screen. You can estimate the difference in path length of the ghosts from seeing what percentage of the screen they are delayed. In some cases, the strongest signal is not the most direct path, so the TV will sync on it and you'll see "pre-echos" from the more direct path(s) that "wrap-around" to the previous scan line.


Radio waves travel about 1000 ft. per microsecond.


For digital signals, the echoes must be within the delay range covered by the receiver's adaptive equalizer; if they are, it's possible for the receiver to cancel them completely. However, if they're real strong, receivers may still have trouble syncing on the signal in the first place. But in general, a receiver with a longer equalizer will deal with multipath better than one with a shorter equalizer.


Multipath delayed longer than the equalizer length can cause big trouble, unless it is so low in amplitude (~-20 dB compared to main signal) that it does so little damage to the signal that it doesn't need to be cancelled.
 

·
Registered
Joined
·
304 Posts
Discussion Starter · #3 ·
So how many microseconds can receivers deal with?


In my case the large antenna tower is nearly on

the direct signal path. I could estimate the

number of extra feet the reflected path might

involve.


And what type of structures, at what distances,

would reflect relevant amounts of signal?


A smaller object that is closer would have a

similar effect to a larger object that is farther.

In my case the object is large, a large antenna

tower, but several miles away.
 

·
Registered
Joined
·
232 Posts
A friend of mine works in DTV equipment. They have a piece of test

gear that displays the cancellation taps in their equalizer in real time.

He says that you can see airplanes moving though the path because they

change the delay and amplitude of the taps. I believe he said the

chip they use supports up to 500 simultaneous cancellations. I don't

remember the delay window on his stuff, but I looked up another chip, NxtWave Communications' NXT2000 receiver chip, and its window is from

4.5 to 44.5 microseconds. (The PR is at http://www.nxtwavecom.com/pr_prs_042600.html).


Ernie
 

·
Registered
Joined
·
304 Posts
Discussion Starter · #5 ·
Let's calculate:


1000 ft/usec

65 usec to cross the screen horizontally


If a signal went an extra 65000 feet it would arrive a full screen late.


That's roughly 13 miles extra.


The range of the chip mentioned is 0 (-4.5?) to 44.5 usec, which works out to 0 to 44500 feet or roughly up to 9 miles extra travel.


The reflective surfaces inside my home, or even my neighbor's home, are nearly zero on such a scale. If the digital tuners are good at dealing

with short delay multipath, it wouldn't be those structures.


The large antenna tower is a several miles away, but at most 1 mile of

multipath delay. Even a wimpy tuner should compensate?


There is a 600 foot high chunk of exposed granite, maybe half a mile in diameter, smoothed over like a giant pebble. It's maybe 10 miles away, and not at all on the path, so it would create a 10 or 15 mile multipath delay. It sticks up like a sore thumb. This would go over the ability of even a good tuner to compensate for? Still, at that distance I don't see how it could yield a strong multipath signal, whatever the delay.


In the end, I've switched from an omnidirectional to a yagi, pointed it at the broadcasters (and the unrelated large antenna tower), and lock on all the time to all 4 available channels. I'm losing interest, fast.


 
1 - 5 of 5 Posts
Status
Not open for further replies.
Top