[Ham-80211] "Commodity Class Phased Array Antennas" for HSMM?

[Gerry Creager N5JXS]

David Young wrote:
> On Sun, Aug 01, 2004 at 06:50:24AM -0500, Gerry Creager N5JXS wrote:
> 
>>While the design of the antenna element only has to be done once, its 
>>implementation has to be done many times, as does the interconnect.  The 
>>steering logic, which drives the elements in a manner to derive 
>>horizontal and vertical tilt, and gain, has to be designed and built. 
>>The phasing array or harness has to be fabricated.
> 
> 
> I'm not sure what you mean by the interconnect.  I cannot speak about
> RF interconnects.  However, for the control interconnecct, I think that
> you could use a PC's audio output.  I'm not kidding.  Audio outputs move
> voice coils all the time....

The interconnection I spoke of is the combination of stripline and 
transmission line that retains the characteristic impedence and 
introduces the appropriate phase delay to make these things steer. 
Since I couldn't see the images in the PPT (didn't get them, or 
OpenOffice didn't like 'em, don't know which) I couldn't completely 
evaluate the idea.

If you're physically steering these things, I'm really confused.  I 
think it'd be useful, though.

> I am confident, as a "software guy," that the software problems are
> tractable.  The hardware problems look pretty tough to me.  A software
> feedback loop based on received signal strength could "train" the beam
> onto the station that you want to talk to by moving the voice coil.

The hardware part is tedious and potentially expensive because of 
element yield, but not particularly hard.

> Regarding fabrication, my thinking is that you would download an "open
> source" PCB design from the web and send it for manufacture.  (Although,
> there are electrical engineers who make PCBs at home.)  You can buy a
> 6"x6" piece of alumina for less than $20 on-line.  I figure machining that
> will be a pain, but it is nothing a determined HSMM hobbyist could not do.

It may be beyond my machining capabilities:=)

> BTW, the steering is horizontal only, although Kajiya tells how you could
> do vertical tilt.  Horizontal steering is perfectly adequate for a lot of
> applications.  For example, my main interest is in mitigating 2.4GHz
> band interference while keeping a "mesh" neighborhood network connected.
> You can steer the beam away from most interference with a horizontal rotation.

Likely, with a little programming, you can train the system to steer 
toward the peak of the desired signal, or use the hueristics to optimize 
the desired and interfering signals...

> Kajiya proposed to build a 4-sector system with one of these antennas
> in each sector, but that requires RF switches.  A 2-sector system might
> just work, and no RF switches are necessary if you re-use the antenna
> diversity switch.

PIN diodes don't work too well at these frequencies, but switches are 
do-able.

I'll try to find a working, non-infected Windows system this week and 
look at the site again, and try to understand his thoughts.  If there'd 
been an e-mail add for him, I'd have already e-mailed Kajiya for some 
more detail...

Gerry
-- 
Gerry Creager -- gerry.creager at tamu.edu
Network Engineering -- AATLT, Texas A&M University	
Cell: 979.229.5301 Office: 979.458.4020 FAX: 979.847.8578
Page: 979.228.0173
Office: 903A Eller Bldg, TAMU, College Station, TX 77843