276 



EXPERIMENTAL PROCEDURES 



during the interval of the pulse. Thus, an accurate 

 record of the ping length appears on the film. 



With this system, the minimum recordable rever- 

 beration signal was limited by amplifier noise during 

 calm water conditions, and by water noise when the 

 sea was choppy. 



Equipment C 



In the early part of 1943, new equipment was put 

 into operation by the UCDWR' Reverberation 

 Group. This equipment can be used with a wide vari- 

 ety of transducers and was originally provided with 

 four distinct frequency channels — 10, 20, 40, and 80 

 kc. However, any four frequencies between 10 and 80 

 kc could be used in the projector circuit by proper ad- 

 justment of the oscillator resonant circuit.* Receiver 

 circuit changes to accommodate different frequencies, 

 such as provision of properly tuned input trans- 

 formers and band-pass filters, could also be made 

 easily. It appears from later UCDWR memoranda ^'^ 

 that this system was altered to include a 24-kc chan- 

 nel and that this channel has actually been used in 

 the majority of the reverberation runs made with this 

 system. 





lOOMS PING 



Figure 4. Shape of signals sent out by equipment C 

 of text. 



The power output with this setup varies with the 

 transducer employed. With the JK transducer at 

 24 kc, the power output averages about 100 db 

 above 1 dyne per sq cm.* A block diagram for this 

 system, assuming a single transducer unit, is given 

 in Figure 3. 



This system differs from systems A and B in a 

 number of respects. A major innovation was the use 

 of electronic timing circuits to control the ping length 

 and keying interval, instead of the complicated me- 

 chanical motor-driven schemes described previously. 

 The changeover relay circuit, used with a single pro- 

 jector-receiver, is also electronically timed, as are the 

 step attenuators which vary the gain in the receiver 



circuit. The pulse projected by this system is practi- 

 cally square-topped. Figure 4 gives photographs of 

 the signal shape for signals of 10 and 100 msec. 



The receiver circuit was specially designed for 

 stability in operation. Positions of the gain changes 

 were automatically marked on the film by means of a 

 flashing lamp. It is clear from Figure 5 that the 

 transients during gain changes are not marked 

 enough to be troublesome. In this illustration the 

 timing trace can be seen at the top of the film; the 

 positions of the gain changes are indicated by spots 

 on the film below the oscillograph trace, which pre- 

 cede the actual gain changes by a fixed distance on 

 the record. 



Because of its greater convenience, its suitability 

 for a large number of transducers, its elimination of 

 transients, and the square-topped shape of its emitted 

 signal, this system is in many respects a considerable 

 improvement over systems A and B. 



Equipment D 



This system is described in references 3 and 7. The 

 projector used with this equipment, the EBI-1, 

 generated a pressure at 1 yd, on the axis, of 104 db 

 above 1 dyne per sq cm. The receiver was a pre- 

 liminary model, and had a number of faults.* Many 

 of the basic features of this equipment are similar to 

 those in systems A, B, and C. However, the detecting 

 mechanism used was not a cathode-ray oscillograph, 

 but a Miller galvanometer mounted in a modified 

 oscillograph camera. Because the galvanometer could 

 not follow 24-kc vibrations, it was operated at 1,000 c 

 by heterodyning the received reverberation to this 

 frequency. One difficulty with this system is that a 

 small percentage change in the i-f oscillator frequency 

 (rated at 251 kc) caused a large deviation in the out- 

 put frequency from 1,000 c, thereby introducing an 

 error since the response of the recording galvanometer 

 is not wholly independent of frequency. In this sys- 

 tem, another galvanometer element was used to 

 record the current fed to the transducer during each 

 ping, while a third galvanometer element was used to 

 make the timing marks. 



The Miller galvanometer is naturally resonant at 

 2,500 c because of its mechanical inertia, and there- 

 fore cannot follow the variations in reverberation 

 intensity with the detail possible with the cathode- 

 ray oscilloscope used in systems A, B, and C. How- 

 ever, the Miller galvanometer is convenient to use 

 and is certainly capable of following the variations in 

 reverberation intensity with sufficient detail for the 



