On April
4, 2002 we began the extension the network to Machipongo
Station at the north end of Hog Island using 802.11b radios for both
cloud and uplink. Local Nature Conservancy employees Richard Ayers
and Marcus Killmon joined LTER researcher Philip Smith on this trip,
helping out with site preparation, power systems, and installation.
Uplink
to Broadwater
We used
a Teletronics CPE (Customer premises equipment -- the client
radio to link to Broadwater Tower's Zcomax access point) with an
Orinoco Silver card for an 802.11b interface.
Teletronics outdoor-ready CPE
Since
it was going to be an unamplified link, we used a 24 dBi
Radiall/Larsen 0.6 meter solid dish to make the 9 kilometer link back
to Broadwater Tower. As mentioned in Diary 44, this would allow us
to
use the full 2 megabit bandwidth of Broadwater tower's Wi-Lan radio
for the backhaul to Oyster, rather than use half of it for relaying
from Broadwater to Machipongo Station.
In retrospect,
we could probably have used a less expensive (and less
expansive) antenna for the uplink to Broadwater Tower; tests showed
that you could make the link acceptably using a 14 dBi panel antenna
costing $60 (versus almost $400 for the 0.6 meter dish).
Large Dish, Small Panel
Problems
Although
the uplink worked flawlessly (and continues to this day to do
so), the access point was another story. Our laptop was unable to
connect through the steel ceiling of the observation tower atop
Machipongo Station. Thinking this to be caused by the metal, or
perhaps by the toroidal propagation pattern of the antenna, we took
our tests outside. Unable to get a link outside either, we realized
that there was a real problem.
Further
tests showed that the signal was better without the amplifier
than with it. So we took radio, amplifier, and all back to Colorado
to test in the lab.
Before
leaving we set up a webcam, connected by Cat 5 Ethernet to the Teletronics
CPE unit, to see how well the link to Machipongo Station worked (with
or without a wireless Internet cloud at the north end of the island).
It worked fine, and continues to do so as of this
writing.
Back
at the Old Colorado City office/lab, we tested various components
of the Machipongo Station access point -- amplifier, radio, cables,
antenna -- and concluded that the problem was due to the low broadcast
power of the Zcomax access point. This model "XI 1000" had
a nominal output power level of 20 milliwatts (mW).
Why does
that matter to an amplifier? It's because the RF amplifier
(this one made by Young Designs, but it's true of most such units)
is
bi-directional. That is, it amplifies both incoming and outgoing
signals -- just not at the same time.
Normally,
the amplifier is in receive mode, boosting what comes in
over the antenna by 16 dB, or a 40-fold increase in power. When the
amplifier detects a signal coming from the radio, the amplification
direction is switched from receive to transmit, and the outgoing signal
is boosted 14 dB or 25-fold. In the case of this amplifier, the manual
said that 3 mW should suffice to move the amp into transmit mode.
The Zcomax
transmit power, even over a short cable, was apparently
insufficient to get the YDI amplifier to switch to transmit mode.
This was doubly interesting because an "identical" Zcomax
radio at
Broadwater Tower was apparently driving a teletronics amp well enough.
We had intentionally used different amplifier brands at Broadwater
and Machipongo for performance comparisons.
(Note:
In the two months since this was written, but prior to web
publication, we discovered that the amplifier we used as a control
in
the lab, was in fact itself malfunctioning. Sigh. We'll retain the
above text, though, to show the steps taken in problem solving.)
We returned
to Hog Island with the Teletronics amplifier, which had
worked in Colorado. Perhaps it was the altitude or humidity, but the
Teletronics amp failed at Machipongo Station.
As a
temporary fix, we rearranged the cabling so the distance between
the radio and the amplifier was minimized, at the expense of
increasing the distance between the amplifier and the antenna.
Incorrect wiring used to drive the amplifier
This
is not the proper way to set up an amplifier. The amp
should be as near the antenna as possible, to minimize noise ingress.
Diary 47 will recount the final installation.
-Tom
Williams
Special Assistant
NSF Wireless Field Tests
