UNDERWATER WEATHER STATION 371 



the canyon is not understood, although the volume of material transported has been measured. 

 Extensive surveys of the canyon head (3) indicate that about 200,000 cubic yards of sand is lost 

 into the canyon each year. A series of current measurements have been made over periods of 

 days and weeks at several stations within the canyon. One of these stations is within scuba div- 

 ing depth at 150 ft at the head of the canyon (Fig. 136). Currents have also been measured on 

 the canyon floor just below the Sealab site at a depth of 650 ft. These measurements were ob- 

 tained by sending a self-contained instrument package down a taut -wire mooring to the canyon 

 floor (Fig. 137). After a predetermined length of time weights are released, and the instrument 

 package returns to the surface, where it is retrieved by scuba divers. The 650-ft station was 

 not occupied during the Sealab measurements because of fouling by the mooring lines from the 

 surface support vessel, Berkone. However, a taut-wire mooring at a depth of 215 ft just north- 

 west of the Sealab II site was instrumented during the Sealab operation (Fig. 138). There data 

 were compared with data from the underwater weather station which was at a similar depth on 

 the rim northeast Sealab II. 



INSTRUMENTATION 



Data from all sensors except the two current meters at the bottom of the canyon were 

 transmitted to a control center (called 'TDenthic control") at the shoreward end of the pier, 

 where it was recorded in both digital and analog form. Pier-end data were transmitted by di- 

 rect cable, but data from the weather station were transmitted through a telemetry system 

 which had its seaward terminal in an underwater benthic chamber. Six 2-conductor cables and 

 three 4-conductor cables connected the instrument array on the weather station platform to an 

 equipment rack inside Sealab. The equipment rack furnished power for the sensors and con- 

 ditioned the signals from the sensors for transmission via the telemetry system. It was es- 

 sential that the variable -resistance sensors receive a constant excitation current. It would 

 have been impractical to furnish an individually regulated constant-current supply for each 

 sensor, so a single 300-volt constant- voltage supply was installed. Individual 300,000-ohm re- 

 sisters connected to the 300-volt supply furnished an excitation current of one millampere for 

 each sensor. Since the resistors were mounted in Sealab, the amperage available in the water 

 was very low. A 12-volt power supply, also mounted in the equipment rack, furnished regulated, 

 constant -voltage power at about 3/4 ampere for electronics packages incorporated in the cur- 

 rent members and the Vibrotron pressure sensor and for signal power amplifiers in the equip- 

 ment rack. 



An analog-to-digital converter inside Sealab changed the signals from variable-resistance 

 sensors to digital form to be transmitted via the telemetry system. Signals from the current 

 meters and the Vibrotron needed no transformation. All sensors except the Vibrotron could be 

 monitored inside Sealab with Rustrak chart recorders. Digital channels of 12 bits each were 

 sampled every 6 or 12 seconds by the analog -to -digital converter. Some of the more important 

 data channels, including the current and pressure sensors, were connected directly to the telem- 

 etry system and could be sampled as often as desired (within limits imposed by the nature of the 

 signals). However, analog signals from the variable-resistance sensors could be telemetered 

 with any great accuracy due to drift in the telemetry channels. 



All signals were connected from Sealab to the underwater benthic telemetry chamber via a 

 multiconductor cable. From here they were transmitted to the shore station via an amplitude- 

 modulated, multichannel carrier telemetry system on a single coaxial cable. 



Savonius Current Meter 



Accurate measurement and recording of low-period, low-velocity undersea currents 

 prompted a modification of the reliable and time-tested Savonius rotor (Fig. 139). A miniature 

 model, designed by Mr. J. M. Snodgrass of the Scripps Institution of Oceanography, was con- 

 structed of "Cycolac" plastic sheet and balanced to be neutrally buoyant in sea water. The 

 rotor was mounted on bearings of sapphire and tungsten carbide. Sixty equally spaced holes 

 near the periphery of one rotor end plate interrupted a light beam as the rotor turned, produc- 

 ing 120 electrical pulses for each revolution. One pulse was generated as the beam passed 

 through each hole, and another pulse was generated as the beam was interrupted. This pulsing 

 output signal has proven to be most reliable in transmitting data over long telemetry channels 

 because its information is relatively immune to amplitude modulation caused by normal noise 

 pickup during transmission. 



