Power,
Power, Power
In the
past year it has become increasingly apparent that the biggest
problems in this business are not caused by radio distances, antennas,
Ethernet converters, serial ports, and the like. As the title
foreshadows, the biggest problems have been related to the acquisition
and maintenance of reliable electrical power in the field.
Broadwater
Tower
Broadwater
Tower is outfitted with a proliferation of devices: a
Wi-LAN radio to provide an Internet backhaul to Oyster, VA; a
Zcomax
radio to provide an 802.11b wireless cloud for the southern end
of Hog
Island; a Teletronics amplifier to boost the Zcomax radio's output
power and input sensitivity; a Panasonic camera (with pan/tilt/zoom
capability) to view the island and surrounding waters; an Axis
web
server to distribute images coming from the Axis camera; and an
Ethernet hub to tie them all together.
These devices
do not all run on 12 volts. The Zcomax does. The
WiLan uses a power supply output of +12, -12, and 5 volts. The
little
hub runs on 3 (or is it 5?) volts. Axis's web site says their
camera
runs on 12 Volts AC (You read right -- Alternating Current). And
so
on.
We learned
that it was possible to use resistors, little solid
state devices, etc., to create a multi-tap, multi-voltage power
supply. But such an arrangement would be highly customized, and
not very replicable. And the purpose of this project is to provide
viable
models for field scientists.
So we took
the easy-but-inefficient route, and used an inverter, a
power strip, and a large number of external power supplies. The
inverter gets its power from a pair of deep-cycle batteries, which
in
turn get their power from an array of six 50-watt solar panels.
Broadwater Tower's Solar Panels
This arrangement
was not sufficient to provide 24x7 power for the
attached equipment, at least not through an inverter and a bunch
of
power supplies and transformers. After about a week of continuous
operation, the batteries had drained to the point that the inverter
automatically shut off. Compounding the problem, this particular
inverter model does not turn back on after its power source
turns off, even after power is restored.
As a stopgap
solution we set up an inexpensive appliance timer that
shuts down everything but the inverter, which would not be kind
enough to turn back on later, from roughly dusk to dawn. A more
permanent solution, not yet implemented as of this writing, has
been offered by Nature Conservancy staff, who have offered to
re-engineer the solar/battery combination, in the process more
than doubling the
capacity by the addition of at least 6 more 50-watt solar panels
to
the mix.
Further
Issues at Broadwater
We found
that the Axis server tended to spontaneously shut down an
hour or so into each day's operation. Guessing that it was caused
by
the heat from all those power supplies, we separated power from
"functional" equipment by moving one inverter and six
power supplies
to a separate box.
Power supplies are in the blue box on the left.
Even if
it works, the goal of replicability is in question. It is
not necessarily reasonable to ask a researcher (who presumably
just
wants to gather some data without requiring someone to hike to
the
site every couple of weeks) to set up a dozen or more 50 watt
solar
panels.
Machipongo Station
At Machipongo
Station we had a somewhat easier time, due to three
factors:
