274 



EXPERIMENTAL PROCEDURES 



keying relay. The ultimate keying control was a 

 synchronous motor-driven pair of shafts. Disks 

 affixed to these shafts operated microswitches which, 

 in turn, controlled the circuit containing the keying 

 relay. By adjusting these disks, it was possible to 

 choose any ping length between zero and several 

 hundred milliseconds, and to control the interval 

 between pings. 



Because reverberation invariably decreases rapidly 

 with time, the receiving system must be specially de- 

 signed to handle a wide voltage range. For this pur- 

 pose, a variable resistor was built into the receiving 

 preamplifier, the amount of gain being controlled by 

 relays. By a keying arrangement similar to that for 

 controlling the ping length, the equipment could be 

 adjusted so that a predetermined amount of resist- 

 ance could be removed from the receiving circuit at 

 any desired time after midsignal. Thus, as the rever- 

 beration intensity decreased, the gain of the pre- 

 amplifier was increased in steps. 



The output voltage from the receiving amplifiers 

 was fed directly to the plates of the horizontal deflec- 

 tion circuit in a Du Mont Model 175A oscilloscope 

 using a short-persistence screen; the vertical deflec- 

 tion circuit was not connected. A continuous record 

 of the oscilloscope deflections was obtained by the 

 use of a fixed optical system and a camera with 

 moving film. Since the spot on the screen moved 

 horizontally, the film moved vertically downward, 

 taking some time to come up to the desired speed of 

 12.5 in. per sec. Because the film speed at a given 

 instant was thus not known accurately, an accurate 

 timing record of some sort had to be photographed 

 along with the oscilloscope reflection. The chosen 

 type consisted of the successive images of a slit which 

 were illuminated by the short-duration flashes of a 

 strobotron tube driven by an electrically controlled 

 fork. 



When operating properly, this system makes a 

 faithful record of all intensity changes in the received 

 reverberation, since the cathode-ray oscilloscope suf- 

 fers from no mechanical inertia effects. However, 

 since it is impractical to run the camera at speeds 

 rapid enough to resolve individual cycles, only the 

 time variation of peak reverberation intensity is dis- 

 cernible on the record. Thus this equipment cannot 

 be used to determine the frequency characteristics of 

 the reverberation. 



In practice, certain difficulties were experienced 

 with this system. For example, when the projector 

 and receiver were close to each other, difficulty was 



experienced because of blocking of the receiver ampli- 

 fier by the received ping, during the period when the 

 projector is radiating its sound pulse. If this blocking 

 is not eliminated, it leads to what may be described 

 as a period of paralysis which lasts for a time after 

 the end of the ping. During this period of paralysis 

 there is serious distortion of the amplitude of the 

 received reverberation. Another problem, also in- 

 volving blocking effects, was the elimination of 

 transients originating during the keying-in of gain 

 changes. These transients could not be entirely elimi- 

 nated in the final versions of this equipment. 



In the derivation of the theoretical formulas for 

 reverberation in Chapter 12, it was assumed that the 

 projected signal was "square-topped," or, in other 

 words, maintained a constant amplitude for a definite 

 interval. No actual ping has this ideal rectangular 

 shape, since some time is always required for the ping 

 to build up and die away. Figure 2 illustrates the 



Figure 2. Shape of 13 MS ping from QCH-3 pro- 

 jector. 



shape of a 13-msec ping sent out by a QCH-3 pro- 

 jector with electronic setup A and recorded by a 

 system with flat frequency response. It will be noted 

 that a definite time, about 1 or 2 msec, elapses before 

 the signal reaches its maximum value. This maximum 

 value is the same as the steady state level for a signal 

 of indefinite duration, but is not held for long; 7 msec 

 from the start of signal emission, the signal level in 

 Figure 2 has diminished below its maximum value by 

 about 4 db. After 13 msec, the signal dies away; how- 

 ever, the rate of decay is measurable. 



Other photographs were taken of longer signals, 

 more than 100 msec in length. In all these, the signal 

 attained its maximum in 1 or 2 msec, fell to 3 or 4 db 

 below maximum at 7 msec, and held a fairly steady 

 level 3 to 4 db below maximum between 7 and 50 



