845 
9. From Plates 7, 8, 9 and 10 it can be seen that the amount of 
the initial overshoot (or "undershoot") of the step-transient response 
depends on the value of the terminating resistance Ro employed with a 
constant value of terminating capacity Co. The amount of this initial 
overshoot (or "undershoot") is particularly important in the faithful 
recording of an applied step signal and hence also in the faithful recording 
of an applied exponentially decaying signal having an initial rise time 
approaching zero. For example in Plate 8 for m = 2, the calculated initial 
value of the relative amplitude for k = O is 3.00 and therefore does not 
appear on the scale of Plate 8. This can be called an overshoot of 200 
percent. As the value of k is increased for m constant, this overshoot 
decreases, becomes zero for a particular value of k, and then becomes an 
Soya eas Plate 8 indicates an initial overshoot of 20 percent for 
= 1.5, 9 percent for k = 1.75, zero percent for k = 2, and an initial 
wundezehoote of 25 percent for k =» 3 (corresponding to a relative amplitude 
of 0.75 for k = 3). Similar results are plotted in Plates 7, 9, and 10. 
From Eqs. (4) and (2) it can be shown that there will be zero 
overshoot (and also zero "undershoot") whenever k and m are selected to 
satisfy the condition. 
(5) k= 24m 
10. Plates 7, 8, 9 and 10 also indicate the magnitude of the oscilla- 
tions existing for time values up to t = 5RC. In general the oscillations 
are less in magnitude for those values of k and m which result in little or 
no overshoot. Hence for the optimum step-transient response for a given 
value of m, it is desirable to select a value of k which will yield little or 
no overshoot and therefore a reduced magnitude of oscillations. 
Step-Transient Response Oscillograms 
11. An experimental verification of the results indicated by the 
calculated step-transient response curves of. Pletes 7, 8, 9 and 10 was 
obtained by an oscillographic study using the same 955 ft length of special 
Simplex twin-coaxial cable as was used for the steady-state measurements. 
Typical oscillograms for m = 2 and various values of k are shown in Plate ll. 
To obtain the oscillograms, a 10 ke sine wave from a Hewlett 
Packard Model 200 C Audio Oscillator was supplied to a Hewlett Packard Model 
210A Square Wave Generator which was operated single ended. Since the 
output impedance of the square wave generator was 500 ohms from one side to 
ground, a 50 ohm resistor was connected across the output terminais s0 that 
the test wave form would be obtained from a low-impedance source. The output 
of the square wave generator was applied to the cable through an input 
capacity C, of 100 micromicrofarads. One coaxial line of the 955 ft Simplex 
twin coaxial cable was used and was terminated with the compensation network 
of Plate 1, using various values of k and m. The output voltage V, of the 
aa, ee NOIM 10467 
