[NCARC] 2.4 Ghz 802.11b info

[James Cizek]

   By the number of responses I got back from that last post
   on 2.4 Ghz stuff, I figured I'd throw this back to the list for everyone.

   My experimentation with 2.4 Ghz 802.11b stuff with was off the shelf
   cards made by D-Link.  I used 2 PCI cards that had external antenna
   jacks on the card.  One of the tricky things about 802.11 stuff is
   the connectors.  Everything is reverse polarity.  The card had a 
   reverse polarity SMA connector.  Some other antennas I had around
   (original 802.11 commercial panel antennas by Cisco) have reverse polarity
   TNC.  This is a good place to go for all your "weird" connectors:
   http://www.hyperlinktech.com/

   Now for antennas:
   The more gain and directivity you have, the better.  I tried using
   yagi antennas both horizontal and vertical polarization.  Most of these
   attempts failed badly.  There is just too much noise for this to work
   correctly over a large difference when you are working with such
   a low signal.  At a distance of 2 miles, the "client" side was hearing
   the signal from the "host" and could be seen on the signal meter
   but no link could be established.  This was using 33dbi gain yagis.

   So, scrapping the yagi idea, I went to the dish.  There is little to
   say about the dish except that it works, and works well.  I used 2 18"
   RCA directv type DSS dishes.  I just hacked the lnb off and mounted
   my antenna where it's focus used to be.  (Focus does not change
   with freq). 
   I used a dipole feed at first.  This worked fine in either polarization
   over a 15 mile range (longest we tested).  But I wanted more signal
   as it was a little weak.  It worked but was a little weak.
   In summary on feeds,  there are 2 that work fantastic.  a 5.25 turn
   helix (cirular polarization) and a splash plate feed (circular also).
   I found that going to circular gained us almost 3 db in every case.
   Not due to increased signal, but in decreased noise.  The majority
   of "stuff" out there is either horizontal or vertical only. Since
   you will have a 3db difference between either linear polarization and
   circular, it makes sense.

   Finally, the cable.  Use hardline and N-connectors period.  Unless
   your run between the net card and the feed is less than 3 feet,
   your loss in the cable will be so bad your system will be deaf.
   I used 3/4" hardline from the windowsill to the top of the tower (30 foot
   tower) and used 1/4 flexible hardline from the top of the tower to the
   dish feed, and 1/4 flex from the window sill (the cable was strung through 
   the window during the test) to the PC where a small jumper was created
   (about 6 inches long) to convert from N connector to reverse SMA.


   One more thing I have to mention:

   Amsat-Oscar 40 uses 2401.323 as the middle of it's primary downlink.
   This transponder on the satellite is right in middle of 802.11b
   land at the lower edge of channels.  Although it doesn't seem to be
   a huge issue as 802.11 is earth based and your antennas are usually
   pointed high while using ao-40, bringing up a link on channel
   0 or channel 1 will still blast the downconverters.  I had some
   neighbors that put up a wireless in their house that severely
   effected my link.  I asked them to change it to channel 11 and that
   solved all the problems.  Just an FYI because there are a few people
   around town using these frequencies in a "weak-signal" fashion !!

   Here is a link to a very small perl program I wrote to quickly
   estimate dish gain, bandwidth, and focus.  I find this really
   useful when comparing dishes to see how much extra you really
   are getting when going to a bigger dish and how hard it will be
   to aim it!
   http://www.jamescizek.com/dish.html

   Please note:  This program is not entirely accurate for 18 inch
   DSS type dishes.  The focus will be off a little.  This is due
   to the fact that most of the DSS dishes are not parabolas, they
   are a section of a parabola.  This means that in order to calculate
   the true focus, you need to calculate the size of the full parabola 
   and enter those numbers.  Then the focus measurement will be on. 
   The gain and beamwidth remain accurate.
   Also remember that offset fed dishes are not "looking" where they
   look like they are "looking" (Is that a sentence?)  
   The actual beam from the DSS dishes is approx 23 degrees higher
   than where the dish "looks" like it is pointing. 
   
   Sorry, that's all that's on the website now. It just got created
   a few days ago and I am working on the content now.  Hopefully 
   I will have some pictures of all this stuff up there soon.
    Hope that helps a little!
    -James    KI0KN
    james@colostate.edu
    ki0kn@amsat.org