More to the point, a 2 input phasing box can only null one source.
I suppose one could get a whole stack of boxes and connect them.
Then the problem becomes "adjusting the knobs", which is a solvable one (at
least in terms of algorithms) - it's the same as you'd do in DSP.
And, early work on adaptive arrays did essentially that: servo motors driving
adjustable reactive components - I'll have to look for the picture of the array
they built at JPL in the 1960s - A whole rack full of gears, motors, and Ls and
Cs.
Other schemes use a vector modulator (which works quite nicely today, because
you can calibrate the system, as long as you've got enough bits in the DAC) - I
used this technique to cancel the Transmit signal into a Receiver in a CWFM
radar at 3 GHz, using a pretty grungy IQ mixer driven from a Teensy.
Even some Adaptive cancellers used at HF in mobile applications in the 90s
still used electromechanical mechanisms - If you've got more than one HF 100W
transmitter on a vehicle, and you don't want to blow up the receivers, you need
something.
Check out something like patent #5152010 to Ashok Talwar at American Nucleonics
Corp (they used to be down the street from where I used to live, and a friend
worked there) HIGHLY DIRECTIVE RADIO RECEIVER EMPLOYING RELATIVELY SMALL
ANTENNAS
Or 4952193, which shows 4 antennas and 4 cancellation paths. INTERFERENCE
CANCELLING SYSTEM AND МЕТHOD
Or even better 4466131 from 1984 - Automatic Separation System
"The present invention is a method and apparatus for separating a plurality of
radio frequency signals, comprising N number of signal collectors, each signal
collector receiving the plurality of signals from a location different than the
other signal collectors, and N being an integer from 2 to infinity; N number of
transmission means, each transmission means conducting the signals received
from one signal collector; and, N (N-1) number of signal cancellation means,
each signal cancellation means having reference, error input and output lines,
the reference line coupled to one transmission means and the error and output
lines coupled to another transmission means; the signal cancellation means and
transmission means being interconnected so that each transmission means is
coupled to the output and error lines of N-1 signal cancellation means and the
reference line of N-1 other signal cancellation means, and no signal
cancellation means having the same transmission means coupled to their output
and reference line; each signal cancellation means substantially nulling in the
transmission means couled to its output line a predominant signal appearing in
its reference line, thus defining a different predominant signal in the
transmission means coupled to the output line of the signal cancellation means;
thus, the signal cancellation means substantially null out N-1 signals from
each transmission means."
Going back a bit further, the above patent cites a patent from 1972
"U.S. Pat. No. 3,648,176 discloses a signal controller to cancel out unwanted
signals by employing an RF potientiometer directly coupled to a drive motor and
a velocity generator whereby corrections in potentiometer ratios may be made
automatically by driving the motor via a feedback loop."
That patent has cool drawings showing the motor operated pots, and this
delightful thing (something I contemplated building at one time):
Typical delays of 2.4 to 430 nanoseconds are achieved by respective lengths of
coaxial cable for each of the switches as tabulated below:
switch 41 300' RG 55B/U coaxial cable 42
switch 42 100' RG
switch 43 50' RG 55B/U coaxial cable
switch 44 30
45 1.7'
Clearly those folks at American Nucleonics have been doing this kind of thing
for a while..
On Sun, 17 Aug 2025 16:49:10 -0700, Jim Brown <jim@audiosystemsgroup.com> wrote:
On 8/17/2025 2:17 PM, David Gilbert wrote:
> I believe those are passive boxes, not DSP
> (somebody please correct me
> if I am wrong). They work fine, but are pretty expensive.
You are correct on both counts, although there are provisions to add
preamp modules to the inputs. And they are not cheap. And what Jim Lux
is saying is that there is some time offset between the two DSP
receivers. We can pretty take what Jim says to the bank -- his day gig
is (was?) as part of the teams that put us on Mars and other cool
places. Jim and I are both contributors to the ARRL Handbook.
73, Jim K9YC
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|