----- Original Message -----
From: "Tom Rauch" <w8ji@contesting.com>
To: "Dudley Chapman" <chief@thechief.com>; <towertalk@contesting.com>
Cc: <richard@karlquist.com>; <jimlux@earthlink.com>
Sent: Sunday, March 06, 2005 3:37 PM
Subject: Re: [TowerTalk] Active phased arrays.
> > With each element being non-resonant and being fed with a
> current source
> > allows for a broadband system and reduces mutual coupling
> between the
> > elements.
>
> Nothing can reduce the mutual coupling between elements. It
> just is what it is for a given spacing and element.
However, with a small non-resonant (actually, the resonant or not isn't as
important as the "small"), you can make the magnitude of the mutual Z pretty
low.
Dudley was also talking about driving the elements with a current source,
which would make the Z immaterial from a radiation pattern standpoint
(although it's certainly important from a designing the current source
compliance range standpoint). It's much like the "quarter wave transmission
line current forcing" scheme. Make the current at the feedpoint what it
needs to be for the desired pattern, and let the rest take care of itself.
In both schemes, there's an efficiency/loss/component limitation concern,
however for Dudley's QRP, that's probably less of an issue. (although, you
might wind up needing kilowatt opamps to actually get a watt into the
element... that's the problem faced by the VLF comm guys.. Many megawatts
into the phased array antenna, all to radiate one watt at <10kHz).
>
> Losses can swamp-out the mutual coupling so you don't notice
> it. I do that with my RX arrays you see at:
> http://www.w8ji.com/small_vertical_arrays.htm
>
> Beverages are easy to phase because the losses swamp-out
> mutual coupling effects.
>
> The problem I see with these large active systems is:
>
> 1.) Ground losses are still the same. You still have the
> same basic element efficiency problems.
That's a universal problem. Nothing to solve it except changing the ground
(sowing the ground with salt and watering heavily?) or moving away from the
ground (buying a tower).
>
> 2.) Mutual impedances cause all the elements to have
> different impedances, and that means the power applied to
> each element for maximum gain and maximum F/B varies wildly
> from element to element with unidirectional arrays. If the
> elements are close spaced and losses are low, you will even
> have NEGATIVE resistance elements. This means multiple power
> amps would produce an overall system having poor efficiency
> both in hardware usage and energy conversion.
Efficiency is in the eye of the beholder, since there's no regulatory
requirement in ham radio for efficiency. So, it really winds up being a
matter of aesthetics and personal preference. Ham Radio, in general, is
probably not the most efficient way to achieve any particular goal (other
than ones specifically ham radio, like DXCC). I know hams who feel that any
real antenna should be built entirely from parts you can buy at Home
Depot. - De gustibus non est disputandum --> "There's no disputing about
taste"
>
> > Naturally, what I just described is too limited to be
> commercially
> > interesting (except for the half dozen 160m/QRP/CW people
> out there), but it
> > would be fun to play with. It also might give me an edge
> for 160m QRP.
> > Since all the beamforming is being done in the DSP, you
> are free to form
> > more than one beam or null at the same time.
>
> The spacing also determines how many nulls or main lobes
> you'll have and where they are at. It is very unlikely
> you'll be able to synthesize patterns with nulls and lobes
> where you want, unless you carefully plan element locations.
>
> When receiving, you could move nulls around at will within
> limits of what the element locations allow, but you'd have
> to live with main lobe locations. You could pick main lobe
> directions, and take what you got (through luck) for null
> locations. When transmitting, you could do the same EXCEPT
> you would almost certainly give up gain and efficiency.
In general, assuming that your elements are suffienciently close (about <
0.5 wavelengths apart), you can form N-1 nulls if you have N elements, at
arbitrary angles. It falls out of basic matrix math, and is the basis of a
whole raft of adaptive beamforming techniques. There are some limitations
on how close two nulls can be, and still be distinguished (related to the
overall extent of the array). Nulls (unlike lobes) can be very, very
narrow. In fact, the real challenge is in making broad nulls with a limited
number of elements (because you don't have enough nulls to can't just stack
up a bunch of narrow nulls next to each other). There are also some
limitations from symmetry (e.g. if all your elements are in a straight line,
you'll get matching/mirrored nulls on both sides of the line).
On TX, who cares where the nulls are.. you want to push the maximum power in
the desired direction. If the elements are "far" apart (i.e. low mutual Z),
then the optimum is generally to phase for time delay matching the spacing
normal to the desired direction. If the elements are close enough together
to couple, you can get superdirectivity, as in a Yagi. There's a limit there
too (superdirectivity means you're storing more energy in the reactive
field, and that increases dissipative losses).
>
> You could be forming a maximum
> > gain lobe in the desired direction while interactively
> nulling strong
> > signals from other directions and doing active noise
> elimination.
>
> It doesn't work that way.
>
> Every element is a source of signal and noise. Both are RF.
> Both are on the same frequency. (If they are NOT on the same
> frequency, then the receiver filter separates them for us.)
>
> All you'd have is a fancy MFJ-1025, and all the same
> limitations that apply to a MFJ-1025 would apply to the DSP
> system.
>
> You can't separate them and process them separately with a
> phasing system. You can't null or subtract noise with
> affecting desired signals from the same direction the same
> way. You can't null strong signals from a given direction
> without also nulling desired signals from the same
> direction. You can't null a signal without creating a
> response change in other directions, and you can't create a
> response peak without creating a null.
Close, but not precisely true. There are three basic situations:
Noise comes from one direction; signal comes from another - adaptive nulling
can completely cancel the noise/interference
noise comes from all directions and is uncorrelated - (i.e. receiver thermal
noise) - can't do a darn thing about it, except you get the usual sqrt(N)
improvement if it's really uncorrelated.
noise comes from all or one direction and is partially correlated; signal
comes from multiple sources (different ionospheric paths/different
polarizations) - you can coherently combine the multiple sources of the
desired signal and reject the common noise/interference signal.
the MFJ does the first one, only.
A power combiner/phasing system can do the second.
Sophisticated DSP can do the third one (and has been demonstrated by
researchers in France and in England on HF skywave paths to great
advantage-- I don't know off hand if it has been done in real time, but
there's no theoretical reason why it can't). The same techniques are used
in high resolution over the horizon HF radar to separate targets from the
clutter around them. If the signal you want has ANY internal correlation
(and this applies to basically ALL ham modulations currently in use, except,
perhaps some well designed spread spectrum techniques), you can use that
information to extract it. (your brain does this for CW fairly well
already)
>
> You are stuck with whatever patterns you can create with the
> element spacing and element locations you have.
But those patterns that you're stuck with cover a pretty darn wide range of
possibilities. It's a MUCH wider range than any single antenna, with however
many elements, on however sophisticated a rotator system, can produce. The
real advantage of the conventional approach is simplicity and the ease of
hoisting it far into the air. And that's a huge advantage in most
applications.. its inherently more reliable, the control interface is simple
to understand, etc.
Jim, W6RMK
_______________________________________________
See: http://www.mscomputer.com for "Self Supporting Towers", "Wireless Weather
Stations", and lot's more. Call Toll Free, 1-800-333-9041 with any questions
and ask for Sherman, W2FLA.
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|