SMOOTH LINES AND SIMULATING NETWORKS 33 



meters P , d, and t which implicitly determine the elements of the 

 excess-simulators in Fig. 7. 



In the first step of this method the two parameters d and P t) are 

 so chosen that the resistance component P of the excess-simulator 

 will be approximately equal to the excess resistance M—k of the line- 

 impedance K, over the specific /-range contemplated, or else will 

 differ therefrom by a nearly constant amount, which can be approxi- 

 mately simulated by a mere series resistance element. In the second 

 step of the method the remaining parameter, /, is so chosen that the 

 reactance component Q of the excess-simulator will be approximately 

 equal to the reactance N of the line impedance, when d and P u have 

 the pair of values already chosen in the first step. The technical 

 procedure in these two steps may now be formulated explicitly as 

 follows : 



First, over the contemplated /-range, plot a set of curves represent- 

 ing P/Po as function of/ with d as parameter; and on the same sheet 

 a set of curves representing (M—k)P {) as function of/ with P as para- 

 meter. To evaluate d and Po choose (by interpolation, if necessary) 

 such P/P -curve and (M—k) P u -curve as most closely coincide. A 

 preliminary idea regarding the useful ranges of d and Po can be readily 

 obtained from the approximate formulas (15-C) and (13-C), together 

 with Fig. 8. 



Second, on another sheet plot as function of/ that particular iV/Po- 

 curve having as parameter the value of P already found in the first 

 step. With this value of P and the corresponding value of d, as 

 found in the first step, plot also a sufficient set of <2/Po-curves as 

 function of / with / as parameter to find the one that coincides most 

 closely with the single iV/Po-curve already plotted. To abridge 

 this step tentative values for / can be readily obtained from the ap- 

 proximate formula (14-C), together with Fig. 8. But the useful 

 range of / can be demarcated more closely by solving tor / the equation 

 obtained by equating the expressions for Q and .V; the value for / 

 thus found is I4 



_ dfN _ d-p 

 l ~ Po 1+dp- 



This may even be plotted, as function of/, to see whether the requisite 

 value of / varies much in the contemplated /-range. 



If the best compromise value of / found in the second step is un- 

 satisfactory as regards simulation of .V by Q, it will be necessary to 

 revert to the first step, choose some other pair of values for d and 



14 It will be recalled that the line-reactance N is practically always negative. 



