The
second phase of our trips to Alaska involved site surveying
and then installing a wireless network to link three Data Loggers
in the Caribou Peak, Poker Flats area, about 40 miles north
of Fairbanks.
Terrain
between University of Fairbanks and Caribou Peak area
In
this case, while the three data loggers are within 3 miles of
each other, the set of them are a long way from Fairbanks and
the University - whose campus is on the far west of Fairbanks.
With the intervening hill masses between, it would be difficult
and costly to try and link the set all the way to Fairbanks.
But the Principle Investigator, Larry Hinzman would like to
be able to fetch the data through the Internet.
So
in one of our previous trips to the area, we learned that the
closest facility to the Caribou Peak area with its own connections
to the University, was the Poker Flats Rocket Range - a NASA/University
facility which launches research rockets into the Aurora Borealis
zones. The on-site director of the facility, Ed Heath, generously
offered to let us set up radios at the Block House - from which
there is a 4 mile line of sight to Caribou Peak to the north.
And we could use the grid power, shelf space for our radio,
and put a Yagi antenna on the roof of the Block House with corresponding
antenna cables.
Before
driving out to the site for the first day's installation work
in a UaF truck driven by Byron Roys, a technical assistant assigned
by Larry Hinzman to support us, and learn how to install such
radios, we picked up larger Solar Panels than we have used before.
85 Watt Panels. We started the work in the Block House.
The
Rocket Range is linked, both voice and data, to the University
by Microwave on towers. And since it is part of the University
we could get IP addresses we need from them.
Rear
entrance to the Block House
Because
the Rocket Range has the Internet connection, we were able to
design the wireless network so that the 'Base Station' Freewave
Radio would be in the Block House plugged into wall power, with
a 15 foot LMR400 antenna cable running up through a conduit
to the roof (which has an earth-grass mat on it to absorb rocket
gasses) and up an already existing antenna pole. We attached
the Yagi antenna, pointed at Caribou Peak to the pole, which
has other small antennas on it.
Inside
the Block House in the Launch Room, we then configured all four
radios, and a novel device from Campbell Scientific, which is
called an NL100. The NL100 is an interface, requiring 12 volts
of power, which can take a serial cable and serial data from
the Base radio in one end, convert it to an IP data stream,
and provide a 10base T Ethernet connection out the other end.
Which device can then be connected by Ethernet cable to the
Ethernet Hub in the Launch Room where all the IP connections
are. By getting a dedicated IP number from the Rocket Range
network administrator then, as far as the world wide Internet
was concerned, its just another IP connection.
Then
the serial cable from the rear of the FreeWave or other serial
radio plugs into the NL100. When further configured with not
only the IP address, but also a 'Port' number (giving a degree
of security to the link), the NL100 then bridges the serial
data flow from the data loggers/radios to the Internet. Anyone
running Campbell's PCW208 3.2 software on a Windows machine
connected to the Internet , when completely configured then
can connect to the data loggers in turn, and download or observe
the data from all the sensors.
Freewave
Radio next to NL100 on top of launch device, close to Ethernet
hub
The first order of business was to configure the four FreeWave
DRG115 Radios to support the following network architecture.
- Master DRG115 and NL100 at the Blockhouse. Wirelessly
linked to the Caribou Peak Slave/Repeater radio 4 miles
away, and through the NL100 to the Internet.
- Slave/Repeater DRG115 on Caribou Peak, which has a
CR10X data logger - named CARIBOUP. Which radio could
both fetch data from the logger there, and relay to either
radio at the sites in the low ground.
- Slave DRG115 at C4MET CR10X data logger site near the
Burn Area down in one valley. About 2 miles from Caribou
Peak.
- Slave DRG115 at CRREL12M CR23X data logger site near
Poker Creek. About 3 miles from Caribou Peak.
A
CR10X data logger connected via a RS232 9 pin cable, to a
9 pin to 25 pin interface , CS/IO to RS232 interface and thence
via 9 pin cable into the Freewave DRG115 radio.
The
PCW208 3.2 software on a test laptop had to be configured at
the 'SOC1' level (Ethernet into NL100) with the serial numbers
of each radio - which act as data logger identifiers (because
the data loggers have no IDs in them) and with special dial
strings that fool the serial wireless network to behave like
a cell phone.
Once
these tasks were done, Mike Willett, Byron Roys, and Tom Williams
set out for Caribou Peak on 4 wheel ATVs while I stayed at the
Block House to run test connections to the data logger through
the laptop software.
One
ATV hauling a trailer with the solar panels and batteries
Mike
and Byron
After
40 minutes of travel up the mountain, the field team installed,
on the large frame holding the Caribou Peak weather station
and data logger, the solar panel, omni antenna, and cables.
On
Caribou Peak
The
Freewave Radio in the Caribou Peak waterproof box with the
CR10X data logger and associated wiring
Success
#1
As
soon as the team completed the connections and attachment, requiring
about an hour and a half, using our Motorola voice radios to
coordinate I connected successfully to the Caribou Peak data
logger via the radios and through the NL100 Ethernet connected
to my laptop and the configured PCW208 software. Both I and
the team on Caribou Peak saw a solid - green carrier detect
light - connection. Which with the clean 4 mile line of sight
path and short antenna cables (which can give 4 dB of loss per
hundred feet) was expected. But it was critical that the solid
link went from Yagi to omni, for the Caribou Peak radio had
to talk to two other radios not in the path of the Yagi directional
antenna feed. Then the team came down the mountain and back
up a shallow valley to the location of the C4MET station. Having
on a previous trip done a handheld radio/antenna check to the
top of Caribou Peak we knew the radios would connect.
It
took most of the rest of the day to install a Yagi high on the
40 foot tower, make the cable with connectors, and waterproof
them, and install the large solar panels, and encase the radio
and battery.
Sensors
atop the tower at C4MET location.
The
base of the C4MET tower. Vertical solar panels
Throughout
the installations in the Poker Flats area, we used separate
solar panels, and separate batteries from those already in place.
There were several reasons for this. First of all, the radio
draws more power than the data loggers. Secondly we did not
want to jeopardize the data logger need for power in the most
severe weather when the site may not be reachable. Thirdly we
wanted to validate Mike's calculation of power need, given the
far north latitude where the sun will be weak low in the sky,
and the temperature can drop below -50 degrees.
We
tested the C4MET radio/data logger link through the relay radio
and everything worked properly.
This
completed the first of two days, with two of the three remote
data loggers accessible from the Internet.
Next
day, we returned to the Rocket Station from Fairbanks and completed
the job on the 3d site.
The
ATVs ready to go again
The
CRREL12M site differed from the others in that it has more sensors
attached, and operates on a newer CR23X Campbell data logger.
Mike
Willett turned over the complete installation at this site to
both Byron Roys and Tom Williams whom he tutored through the
first two installations.
Bryon
Roys up the tower, affixing the Yagi
The
battery box is nothing more than an insulated cooler - heat
from the batteries can help
The
very busy waterproof box for the data logger, the radio, and
all the connectors
It
took a little longer for the less experienced members to complete
the installation, but it worked perfectly when they were done.
Mike Willett, while they were doing the work, narrated some
of the installation details not covered above. That narration
and short videos are at:
Short
Videos
After our entire team returned to the lower 48, I was able to
access all the data loggers through the Internet repeatedly,
showing, at least under the prevailing weather conditions that
we had a robust link to the remote data loggers in central Alaska.
And the data could be accessed as often as desired, interactively,
at any time.
About
a month later, Larry Hinzman's staff was able to take the final
step - interface the PC208W software to a generic Web Server
interface. So anyone in the world with a web browser can now
see the data by accessing the following URL.
http://www.uaf.edu/water/projects/cpcrw/metdata/cpcrw-hourly.html
With
that final step, and only the monitoring of how reliable the
network is under the coming harsh central Alaskan weather conditions,
this Model is complete. It can be replicated anywhere there
are Campbell Data loggers and a Net connection with 1 to 10
or more miles.
