John KD5YI wrote:

"user" wrote in message
...



Take a look here too:
http://cordis.europa.eu/ictresults/i...tures/ID/89327



“The transmit antennas need to be spatially separated by between 10
and 20
wavelengths and a delay applied to one antenna to achieve effective
de-correlation. The amount of separation and delay depends on the type of
environment to be covered,” says Bard.



Can someone explain this de-correlation to me? Is this not the same as a
phase shift?

They're talking about that the received signals from the two antennas
are decorrelated, generally, with respect to the multipath. For the two
direct paths, it is just a time delay difference. But when you throw
reflections into the mix, they'll be decorrelated.

Virtually all DTV receivers use some form of adaptive RAKE receiver to
effectively utilize the energy in the multipath and to provide
resistance to multipath fading. To the receiver, adding another
transmitter just looks like another reflected path.



Would there not be a de-correlation when moving from
broadside to end fire?

Depends.. How far did the phase center move when moving from broadside
to end fire.