The system has three parts: seismic data acquisition; data storage and integrity; and power/heat generation and control. We use off the shelf products for the most part, except for the software of the CPU and the power/heat generation/control subsystem.
The PASSCAL/Reftek DAS (model 72A-08) is a six-channel digitizer that we have set up to record continuously at 1 sample per second (sps) and 40 sps on the three high-resolution channels. We have attached a 3-component geophone to the low-resolution channels.
These parameters are set in the DAS by the software when the
The Anubis CPU is based on an Intel 386 single-board computer running the
The data are stored on two 8GB rotating-platter hard drives. This is one of the more sensitive parts of the system (sensitive mainly to temperature). This component drove the specs for the instrument shelter temperature control. The disks are IDE drives in a removeable sled so that servicing the stations will not involve disconnecting any connectors or unscrewing the drive from its mounting.
The power to the disk drives and to the CPU is controlled by a temperature switch that will shut off all operations if the temperature drops below -5C. At 0C, the Disk Heater board (designed and built by Bruce Long) will draw about 3W and dissipate it in a transistor heat sink in an attempt to maintain warmth close to the hard disks, inside the CPU box. If the temperature nevertheless continues to fall, the disks and CPU are shut off after a one-minute warning signal.
The charge control system was designed and built at Penn State University by Bruce Long under the direction of Sridhar Anandakrishnan. The currents from the two wind generators and the two solar panels are summed. This summed signal is placed across the batteries and the resulting voltage monitored. A temperature sensor attached to the ground lug of one of the batteries is used to determine the float voltage (based on the battery manufacturer specs). As this float voltage is approached and exceeded, the input current is redirected to a set of Heater Plates.
The charge control board also contains the Low Voltage Disconnect (LVD) circuitry. As the battery voltage drops, the loads are gradually shed until all the loads have been shut off. The lowest priority item - the Auxiliary box, which contains the Campbell CR10 and the Argos transmitter - is first shut off. Next the middle-priority items (the reftek DAS and the Anubis CPU) are shut off. Finally the highest priority item - the Guralp seismometer is shut off. At this point the batteries are depleted to about 80% of full charge. Hopefully this level of charge will be sufficient to protect them from freezing, even at the low temps that are possible in Antarctica.
These heater plates consist of three high-current transistors whose heat-sink tabs are bolted to the aluminum plate that forms the base for the batteries. Four of these heater plates (capable of sinking a total of ??20A??) make up the heating system.
As the internal temperature of the battery rises to +10C, the input current is redirected to a set of of four heater plates that are mounted to the aluminum plate that make up the Tower Junction Box outside the shelter. Thus the shelter will not overheat, yet the wind generators will always remain under load, preventing damage.
We use two brands of wind generators. Both produce about 70-100 watts at 12 m/s wind (about 25 knots) and max out at about 170W at 25 m/s (50 knots) and above. Both are capable of withstanding wind speeds of 125 knots. The Rutland 910 is a "furling" model that turns out of the wind to prevent the damage to the rotor. The Ampaire 100 has high impedance internal windings that load the alternator at high currents and prevent it from spinning faster than is safe. As a consequence, the Rutland produces power at lower wind speeds than the Ampaire. The reason we chose to use two different wind generators is so that any systemic or design fault would not shut down the station.
The wind generators are mounted at the top of a 15 foot tower made of 1-1/2 inch schedule 40 pipe. We thread together two sections: a 10 foot lower section and a 5 foot upper section to make the tower. A Tee at the join is used to bring out the power cable. A flange with 4 guy wires is slipped over the tower just above the Tee. We attach these guy wires to deadmen that are buried in the snow.