[Elecraft] Diversity Receive as Noise Canceling?

[David Gilbert]

I'm rather surprised that nobody has noticed the connection between this 
discussion and the one we had a few weeks ago about using the two 
phase-locked receivers in the K3 for directional nulling.  It's really 
the same thing .... feed separate antennas into the two receivers while 
in diversity mode and process the audio to null whatever you want.  It's 
all based upon the fact that the time differential between two RF 
signals is preserved when both are down converted to audio by phase 
locked local oscillators.

And yes, assuming enough remaining processing power in the K3, the 
necessary time shifting and inversion to accomplish a suitable null 
could be done in the K3 (assuming a knob could be highjacked for the 
adjustment), but it doesn't need to be.  It could more practically be 
done in the sound card of almost any computer, and with a lot more 
flexibility.  It just requires two things:

a.  Elecraft to make the phase lock between the two receivers remain 
constant as the frequency is changed.  Ideally the phase differential of 
the local oscillators would be zero ... either forced that way by the 
DSP or by using the same L.O. for both receivers.  If that wasn't 
possible, it would still be usable if the differential were known or 
could be interrogated by a command.

b.  Someone to write a simple bit of software for the computer that took 
the sound card inputs, shifted and processed them, and then feed them 
back out through the sound card to headphones or speakers.

There are several possible uses of such a capability:

1.  Nulling directional noise, assuming two antennas separated 
horizontally.  The software could even automatically search for the 
phase that gave the best reduction in noise.
2.  Nulling an interfering station from one direction, assuming two 
antennas separated horizontally
3.  Nulling an interfering station from a particular arrival angle, 
assuming two antennas separate vertically.  Picture a stack of two 
horizontally polarized antennas (yagis or dipoles), with a loud 
stateside station covering up a weaker DX station.  Presumably (but not 
always) the stateside station  would be arriving at a higher arrival 
angle that might possibly be discriminated against by adjusting the 
phase of the two audio signals to put a notch on it, the same as if it 
were possible to continuously adjust the phase of the two antennas at 
RF.  Granted, ground reflections limit the range of notch adjustment, 
but if the antennas were high enough the concept might be practical, 
although I haven't checked it with EZNEC to see what the limits might be.
4.  Real time monitoring of the azimuth of an arriving signal for 
propagation studies (looking for skew paths, etc).  Just continuously 
rotate the phase shift and look for nulls.
5.  Real time monitoring of the elevation angle of an arriving signal 
(such as the full time beacons) for propagation studies.  Just 
continuously rotate the phase shift and look for nulls.
6.  Real time monitoring of elliptical polarization (both direction of 
the rotation and the relative magnitudes of the two components) ala the 
recent request from KL7AJ.

There are probably others I haven't thought of.

At the time of the prior message thread Elecraft indicated it might be 
possible for them to consider some modifications (either hardware or 
software) in the future that would predictably lock the two receivers, 
but I haven't heard anything more since then.  It seems like too 
valuable a potential capability to just ignore, though.

73,
Dave   AB7E



On 11/30/2010 12:44 AM, Edward R. Cole wrote:
> At 10:06 PM 11/29/2010, you wrote:
>> Noted your comment on noise reduction by phase nulling. Have you
>> looked on info on pages by pa0sim, see www.pa0sim.nl, he has done
>> more or less what I think you are considering. Ie connecting the
>> output from 2 phase locked receivers fed with separate antennas and
>> do the amplitude/phase nulling at AF level.
>>
>> Some of the sound samples on his website are impressive in terms of
>> noise reduction.
>>
>> I would love to see an application which could do this, would be
>> nice to check out if phase nulling works at a site before trying an
>> MFJ unit or similar. I really would like to see this done in the K3
>> as all the hardware is there to do such phasenulling in software,
>> could eg use the VFO B knob for phase control and the volume knobs
>> for amplitude control. Only requirement is to feed the 2 RXs with
>> separate antennas, ie an alternative use of diversity capability.
>>
>> I was not able to respond on the reflector. Would be nice if you
>> could keep the group updated if you get more info on this.
>>
>> 73 de Olaf LA3RK
> Olaf,
>
> Thanks for writing.  I somehow lost your e-mail and wanted to look at
> the pa0sim webpage.  I merely suggested the idea.  I really have no
> prospect for doing this.  I thought maybe Elecraft might be
> interested in the idea; provide firmware to do it.
>
> A modification on doing diversity reception was what I had in
> mind.  Use a small "sniffer" antenna on the aux Rx input to sample
> noise and null it from the main receiver in sw.  I think it would be
> easiest to accomplish in digital before conversion to audio.  All the
> DSP routines are done there:  NB, NR, AGC, APF, DSP bandwidth.  Phase
> shift and amplitude adjustments would be possible in digital, so
> assigning VFO-B would be relatively simple for phase and use the
> sub-Rx volume for amplitude adjustment.
>
> I guess it's now up to Elecraft if they want to explore this.
>
>
>
> 73, Ed - KL7UW, WD2XSH/45