STUDY OF FREE-RANGING SHARKS 467 



latter is superior in voltage and energy density. While superficially appearing 

 to cost several times more than mercury cells, if compared on a cost-per- 

 watt-hour basis, lithium cells appear reasonable and in some cases even less 

 expensive than mercury. One present problem with lithium is the lack of 

 availability of sufficient sizes and shapes, most being large, elongate cells 

 such as AA, C, or D sizes. 



Duty cycle— In a pulsed UST, the duty cycle is normally defined as 

 the percentage of total time that the power-output stage is on, e.g., 10-ms 

 pulses at a rate of 2/s give a 2% duty cycle. Battery life, however, depends 

 on an "average" current drain consisting of a combination of (1) brief, 

 heavy drains during the pulses, often partially "smoothed" by a large capaci- 

 tor across the battery, and (2) continuous low-level drains of those circuit 

 sections that are continuously on. In most transmitters, the first of these 

 factors is the most significant. The obvious strategy of reducing the duty 

 cycle to conserve the battery, unfortunately, can be carried only so far. 

 As discussed earlier, there are limits for both pulse shortness and pulse-rate 

 slowness beyond which signal detection or recognition is impaired. 



Transmitter life— Battery capacity is usually specified as the number 

 of milliampere hours that can be produced before the voltage drops to a 

 certain arbitrary cutoff value, e.g., 0.9 V for mercury cells. Battery life in a 

 UST can therefore be approximated by dividing the capacity (in milliampere- 

 hours) by the average current drain (in milliamperes). It should be realized 

 that, since voltage drops with battery use, the current drawn also drops; this 

 must be considered when calculating average current. Furthermore, the 

 manufacturer's specified capacity is valid only if the battery is used at the 

 specified temperature and constant discharge rate. Under the extremely 

 varying (pulsed) drain conditions of low duty cycle USTs, the specified 

 capacities may or may not be applicable. Thus, the best estimates of trans- 

 mitter life are those based on test runs in which a battery is consumed 

 actually operating the transmitter. Tests on the CSULB shark transmitters 

 at 2-3% duty cycles indicate that actual life is reasonably close to manu- 

 facturer's specified life for 600- and 750-mAh, 8.4-V mercury batteries. 



It is useful to consider how relative changes in range or life affect needed 

 battery capacity. Consider spherical spreading only; to double range, acous- 

 tic output must be raised by 6 dB, and this requires a fourfold (4X) boost 

 in electrical power, thus four times the battery capacity for equivalent life. 

 But since at any significant distance absorption must also be considered, it 

 will actually require much more than a 4X battery increase to double range 

 without reducing life. For instance, Table 7 shows that to boost range 

 from 1 km to 2 km at 40 kHz actually requires a 15-dB signal increase, 

 and this means a 32X battery boost. On the other hand, to double life 

 without changing range requires only a twofold battery increase, to quad- 

 ruple life means a fourfold battery boost, etc. 



Thus, at the usual distances involved in shark telemetry, increased life 

 is relatively easier to achieve than increased range. One can pay dearly in 

 battery drain in attempting to gain "long" range by boosting output power 



