The photomultiplier tubes (RCA-6655A) operate with a negative voltage of 

 700 to 900 V. This 200-V range permits gain adjustment of the tubes so 

 that the responses from all four of the detectors will be equal. The tubes 

 are 2 in. in diameter and have ten stages. The preamplifiers (one for each 

 tube) consist of three emitter-followers in cascade. The preamplifier out- 

 put is matched in a 50-ohm coaxial cable (RG-17^-u) which carries both the 

 positive 2^-V dc power to the preamplifiers and the output voltage pulses 

 from the preamplifiers. These pulses are transmitted to a mixer on board 

 the tow vessel. The single output pulse from the mixer is fed to the amp- 

 lifier of the multichannel analyzer (PIP-400). 



The sides of the cylinder used to transport the detector assembly are fabri- 

 cated with rectangular steel bars to form an open lattice with a minimum of 

 shielding of the detector crystals. This allows the 80-keV eamma and a 



133 



fraction of the JO-keV x-ray from Xe to reach the detectors, which are 

 positioned approximately 2 in. from the surface over which the cylindrical 

 ball travels . The entire device is covered with expanded metal to exclude 

 stones and other debris and to provide mechanical protection (Addendum C-k) . 



The detector housine is weighted with lead to maintain the detectors in a 

 vertical position('see Addendum C-5 f° r description of the device that indi- 

 cates the position of the detector assembly) . At a speed of 3 mph, however, 

 the forward motion of the rolling device causes the detector housing to be 

 ~5 off-center toward the back. Since the count rates are a direct func- 

 tion of geometry, an experiment was designed to determine the difference 

 between the count rates obtained when the detector housing was in the vert- 

 ical position and when it was off-center. The following results were ob- 

 tained for three variations from vertical: 



0° 910 counts/sec 



5° 880 counts /sec 



10° 800 counts/sec 



The difference of JO counts/sec or ~3$> for an angle of 5° (normal operating 

 angle) indicates that the count rate is not appreciably affected by an off- 

 center movement of the detector housing. 



The towing characteristics of the detector assembly were determined by the 

 Naval Ship Research and Development Center, Tests proved 



that with a tow cable 200 ft long 'the detector assembly was stable and re- 

 mained on the bottom at a depth of up to 80 ft at a speed of up to 6.7 fps. 

 Changing the length of the cable will modify the speed at which the assembly 

 can be towed and still remain on the bottom. It was also established that 

 the assembly would remain in the survey position (detectors pointed toward 

 the bottom) 'even when it was pulled rapidly through the water column. How- 

 ever, the tow cable and the plastic tube containing the conductor cables 

 oscillated, and it was evident that, the conductor cables would be damaged 

 if they were towed rapidly through the water for long periods of time. 



C-ll 



