Page:Popular Science Monthly Volume 92.djvu/654

 638

��Popular Science Monthly

��somewhat beyond the point of rest and compress the spring. Thereafter the weight will immediately start downward; and it will continue to oscillate up and down in shorter and shorter strokes until the energy stored in the weight and spring system has been used up. This cor- responds in many ways to the circuit

���Fig. 37: Mechanical oscillating system Fig. 38: Practical air-blast spark-gap

shawn; pulling against the spring until the thread breaks is comparable to charg- ing the condenser until the spark-gap breaks down, and the rapid up-and-down oscillations of the weight and spring are much like the rapid electrical oscillations in the condenser and coil circuit. The spring is analogous to the condenser C and the weight to the inductance coil Lij it is the stress of the spring (and in the condenser) which trips off or breaks down the restraining element (thread or spark- gap), and it is the energy stored in the weight (and in the inductance coil) which carries the oscillations beyond dead center on each swing and so keeps the system vibrating.

Controlling the Oscillation Frequency

The two systems are alike as to another important point, viz., the frequency of the oscillations. We know from experi- ence that the greater the mass of the suspended weight and the greater the flimsiness of the spring, the more slowly the mechanical vibrations of the system. By varying either or both of these we can make the weight bob up and down at almost any frequency we choose. In the same way, the frequency of electrical oscillations in the condenser and coil cir- cuit is almost entirely dependent upon the size of the condenser and coil. The larger the condenser (the greater its

��capacity) and the larger the coil (the greater its inductance), the slower the radio frequency oscillations will be. Thus, by altering the electrical constants of the circuit (e.g., the capacity and inductance), we can make the oscillation frequency almost anything we desire. This matter will be treated in greater detail later.

The next point which should be con- sidered here is the construction of a spark- gap which will work regularly and con- tinuously. Commercial radio practice has brought out a great many types of spark-gap, but years of experience have shown that certain properties must be secured if satisfactory operation is to be expected. In the first place, the gap must always break down at some definite volt- age. It is evident from Fig. 36 that if the potential which established conduc- tivity across the gap varied, the oscilla- tions would begin at different points in each half-cycle and that the oscillation groups would not occur regularly. If the break-down potential were normally 9,500 but sometimes became 8,000, when the lower value held the oscillations would start off too soon in the half-cycle, and the full discharge of the condenser would not be utilized. If it ran up to 11,000 volts, no spark would pass at all, and the charge for that particular half-cycle would be practically wasted in so far as the production of a group of radio fre- quency oscillations was concerned. Uni- formity of sparking potential depends upon keeping the gap cool more than on anything else, since the hotter the gap the lower the potential at which it breaks down. For small powers the necessary cooling may be secured by making the spark-gap terminals large, since then the heat will be carried away rapidly by the mass of metal. For larger powers some form of artificial cooling is used.

A Successful Cooled Spark-Gap

A form of air-cooled gap which has been found satisfactory for many pur- poses, is shown in Fig. 38, and which is largely used by the French. It consists merely of a brass or copper tube forming one electrode and placed endwise to a Hat plate which acts as the other terminal. A blast of air is fed through the tube by way of a rubber hose, and spreads out

�� �