[RFI] Electric company problem

[Jim Brown]

On Sat, 10 Dec 2005 20:51:01 -0500, Tom Rauch wrote:

>The noise
>is sampled, phase inverted, and combined with the noise from
>the main antenna at the same level as that noise. What you
>are really doing is setting the sensitivity of the sense
>antenna and main antenna to the same level at the distance
>and angle of the undesired signal, and inverting the noise
>exactly 180 degrees. 

I agree with almost everything you are saying with respect to 
cancellation of multiple sources and azimuth, except for one VERY 
important thing. A noise canceller is SHIFTING the phase of the 
noise sense antenna so that the resultant signal from the sense 
antenna is 180 degrees out of phase from that same noise on the main 
receiving antenna. That phase shift may be any number of degrees 
from -infinity to +infinity (although -180 to +180 is enough). :) 
The fact that it must be a phase SHIFT and not only a simple 
polarity reversal is what makes the adjustment so tricky (and why it 
must be re-balanced when the frequency changes very much). 

This is a discussion we have had before, and as a practitioner in  
both the audio and RF worlds, I am VERY bitchy about the difference 
between polarity and phase. I know you design these things, and that 
you do that very well, but the way we TALK about how they work is 
very important to all of us understanding them. :) 

Polarity is a two-valued function, and it is not related to 
frequency or time. Phase has an infinite number of values, and it IS 
a function of frequency and time. It is not possible to "invert the 
phase."   

73,

Jim K9YC