On Fri, 27 Nov 2009 11:20:48 -0800 (PST), "
wrote:

Those old schemes where you quickly switch between antennas work well
for voice signals. I've found data channels don't do so well.

A "Homer" works fine for continuous uni-directional data. It doesn't
work well for half duplex such as Wi-Fi. They're really awful for
situations where you hear multiple xmitters with the added bonus of
collisions (i.e. packet data). There are workarounds, but all require
decoding the data first and using a computah to keep track of which
xmitter you're hearing.

I was
told that it has to do with how the data channels are nearly always at
full modulation. One of these days I'm going to hack two radios so
they share one local oscillator and see if I can DF based on phase
comparison of the last IF output.

It will work with one big catch. You'll have to replace the IF
crystal filters with one that has a fairly stable group delay across
the IF bandwidth. If you're working with data that has modulation
components all the way to the edge of the IF filter, you're going to
go through huge variations in phase shift between the carrier and the
filter skirts. Building two matched receivers just adds to the
complexity. OAR (Ocean Applied Research, now part of Cubic) once made
an Adcock antenna DF system that had 3 receivers for measuring the
phase and displaying the direction on a scope.
http://www.cubic.com/cda1/Prod_&_Serv/C4ISR_Prod_&_Sys/DF_Products/df_processors.html
http://www.cubic.com/cda1/pdf/aa1319.pdf

Personally, I think you would do better with multiple remote receivers
and measuring the TDOA (time difference of arrival) at each receiver.
Each burst of RF is time coded and packaged along with signal
strength, GPS phase, and decoded data, and sent off to a central
computah for processing. The time codes are critical as that allows
storage, replay, and post processing. Spewing RF around Area 51 is
probably a bad idea, so post processing make sense. Collect the time
coded data on thumb drives, stuff the data into a computah on your way
out, compute, and replay the captured data later.

Drivel: If you're thinking of doing it with a Doppler ADF, then
please read my previous rants before blundering onward:
http://www.qsl.net/n9zia/doppler_notes1.txt
http://www.qsl.net/n9zia/doppler_notes2.txt
http://802.11junk.com/jeffl/AN-SRD-22/
Those are 15 year old postings about a 35 year old product, but it's
still generally accurate.

Did you ever build and try the 1090 Mhz AMOS/Franklin antenna?

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558