Page:Popular Science Monthly Volume 88.djvu/965

 Popular Science Monthly

��937

��the condenser one at a time, or small additional currents must be introduced by way of the inducti\'e portions of the circuit. These charges must be applied at the instant that the natural (or self- oscillating) charge of the condenser is of their polarity, for otherwise no advantage of increased charge would be had ; similarly, the increments (or additions) of current must be made when the natural current is flowing in the proper direction, for, if not, there would be an opposition to the normal current in the circuit and no increase would be secured. This is as certain as the fact that, in order to make a swing go higher and higher, it must be pushed when it is moving or about to move in the same direction as the applied force; and it is true for the same reason.

Let us now assume that the alternator E in Fig. i is capable of delivering I kilowatt of electrical power at 50,000 cycles per second, but can run safely at speeds as high as that giving 100,000 cycles. Let the inductance and capacity be of such values that the natural frequency of the circuit is 50,000 cycles per second (corresponding to a wave- length of 6,000 meters), and consider that the total resistance is two ohms. If the alternator is started from rest and gradually speeded up, it will

���Fig. 6. A closed and an open circuit (A and B) are used to radiate waves to a receiving antenna, C

produce pulses of alternating voltage at gradually increasing frequencies. These voltage impulses will charge the condenser C first in one direction and then in the other; but very little current will flow, because there is no tendency for these lower-frequency voltages to co-operate by resonance. As the frequency comes close to 50,000 per second, however, the current will corn-

��Fig. 5. Simple antenna circuit

���mence to rise, and at 50,000 cycles it will reach a maximum of about 23 amperes. At this frequency the small voltage additions produced by each cycle of the alternator are impressed upon the condenser exactly in step with the natural oscillation voltages, and the greatest possible oscillation current results. When the frequency is increased beyond 50,000 cy- cles, the resonant value of the circuit, the circuit begins to fall off very rap- idly. If one meas- ures the current at __^ each of a set of fre- quencies near the tuned point, the result may be plotted in the form of a curve like that of Fig. 3, where the intersection over each frequency shows the amount of current indicated by / when the alternator is run at the corresponding speed. It should be noted that the rise and fall are extremely sudden.

Suppose now that this same experi- ment be repeated with all conditions remaining the same, except that the total resistance of the circuit is set at 10 ohms. As the speed of the alternator is increased it is noted that the current begins to rise in the neighborhood of 50,000 cycles, as before, and to fall after that speed is passed; the interest- ing features are, however, that the maximum current is now only 10 amperes, and that the rise and fall near the resonant point are not nearly so sudden as before. By taking a series of careful measurements and plotting them out in curve form, a diagram like that of Fig. 4 may be produced. The slope of the sides of this curve is considerably less than that of Fig. 3 ; the effects of adding resistance have evidently been to decrease the current at resonance, and to make the circuit less sharply dependent upon applied frequency. \Vc know that this means the tuning of the circuit has become less sharp; we know also, that the adding of resistance has increased the damping of the free oscillations in the circuit. These two results are closely related.

Next, the application of these experi- ments to a modern radio telegraph

�� �