MEASUREMENT TECHNIQUES 



The calibrations made on the GILLISS are based on the hypothesis that the 

 voltage induced in an isolated electrode pair is numerically equal to the product of 

 the field strength and the antenna length. Because the impedance of the electrode 

 antenna is very low as compared with the input impedance of the preamplifier, a 

 field-induced voltage at the antenna may be simulated by means of a calibration 

 voltage injected in series with the electrode pair. 



To insure that any changes in systemi sensitivity would be detected, three 

 calibration measurements were made. In one case the preamplifiers were terminated 

 with a dummy resistor at the input terminals, and the calibration measurements 

 were compared with laboratory data taken before sailing. In the second calibration 

 measurement, the dummy resistor was located at the trailed end of the electrode 

 antenna. This second measurement checked the antenna cable and lead-in wires. 

 The third calibration measurement was made on the antenna. The system was 

 calibrated throughout its full frequency range at the beginning and end of the trip, 

 and a spot calibration was recorded during each measuring period. 



Measurements were made on the GILLISS at sunrise, mid-day, sunset, and 

 post-sunset with the electrodes depressed approximately 30 meters below the 

 surface. The ship trailed the antenna at about 3 knots; this slow speed kept the 

 electrode velocity-induced noise to a minimum. 



A power spectrum analysis has been performed on some data from the October 1964 

 trip on the GILLISS. The analysis was performed on 152-second data samples 

 using a 12. 5 cps bandwidth in the spectrum, analyzer. One hundred different fre- 

 quencies were analyzed in the range between 10 and 1000 cps. These results are 

 shown in Figs. 5 and 6. In addition to the power spectra of the electrode antenna, 

 the power spectra of both the dummy antenna and an 8-ohm resistor at the input to 

 the preamplifier terminals are shown. As shown by these two spectra, there was 

 a small amount of hum in the system. Figure 5 compares two samples taken within 

 a half hour of each other. Note the spread of approximately 3 db. Figure 6 compares 

 data of different days — two morning samples and two afternoon samples. Note the 

 possible morning and afternoon differences. 



Ultimately the analysis of these data will result in quantitative ELF electro- 

 magnetic noise levels as a function of frequency, time of day, season, geographic 

 location, and meteorological and oceanographic conditions. 



252 



