190 MEASUREMENT OF PRESSURES 



(2). Resistance and capacitance at receiving end. A better com- 

 pensation developed by Lampson (66) adds a resistance R2 in series 

 with C2, the terminal voltage being taken across C2. Specializing Eq. 

 (5.14) to this case gives for the steady state response 



I'd") 



C ^B^^ (1 + jc^R^C) + C2 COS RoCo^ 



Empirical tests and calculations show that if C2 = 2C the optimum 

 value of R2 is about 1.5Ro and the response for these values is shown in 

 Fig. 5.19 (curve (b)). The resonances are reduced but analysis shows 

 that no combinations of R2 and C2 can eliminate them. The first two 

 terms in the corresponding step response are 



l> R,C l> ZRoC 



f 



This expression, plotted in Fig. 5.20 (curve (b)), shows an improvement 

 over use of capacitance alone but retains the resonant characteristic. 

 It should be noted here that all these results are based on the assump- 

 tion that cable losses are negligible. Bancroft has shown that Lamp- 

 son's network can be adjusted to match the surge impedance of a cable 

 with finite ohmic resistance of the conductors. The appropriate value 

 of capacitance C2 which .accomplishes this turns out to be extremely 

 large (of the order 0.25 /x/), and would give very greatly attenuated 

 voltage Vt. Tests at the Taylor Model Basin have indicated satisfac- 

 tory performance with smaller values/^ but where considerations of 

 sensitivity are a factor it is desirable to have a more efficient network. 



(3). Resistance and capacitance at both ends of cable. A con- 

 sideration of the behavior of Lampson's network, and of cable prop- 

 erties more generally, suggests that the incipient resonances of the single 

 network can be reduced by terminating both ends of the cable, in other 

 words using all the elements in the schematic circuit of Fig. 5.18. The 

 steady state characteristics of this network have been found to give 

 satisfactory results for resistances i^i, R2 about equal to the cable surge 

 impedance Ro, and capacitances Ci, C2 equal to the cable capacitance C. 

 Comparison of calculated frequency response with experimental meas- 

 urements on rubber dielectric cables with surge impedance of approx- 

 imately 50 ohms shows satisfactory agreement (20), considering the 

 approximations of the theory and failure of the cable to have exactly 

 the ideal characteristics ascribed to it. 



^8 These are described in a report by Greenfield and Shapiro (42). 



