466
��Popular Science Monthly
��of any circuit, either from the out- side (by magnetic induction, for in- stance) or internally by a high frequen- cy alternator or other apparatus, a forced alter?iating current of the gene- rating frequency will flow. The fre- quency of this forced current cannot be changed by varying the constants of the circuit, for it is determined by the gene- rating source. The amount of current which is set up for a certain voltage, however, is governed very largely by the circuit constants. As was shown in
��^ . .'. ,
T
��Fig. 4. Direct coupled ^=:/ sender
the January article, the greatest current flows when the applied alternating fre- quency is of the value for which the ca- pacity and inductance of the driven cir- cuit neutralize each other, or that for which the impedance (alternating cur- rent resistance) is therefore the smallest. The other type of alternating cur- rent, called "free," occurs when a con- denser is charged and then allowed to discharge through an inductance and re- sistance (which must not be of too high value). The frequency of this free alternating current so produced is de- pendent entirely upon the constants of the circuit, and, for the same values of capacity and inductance, is practically identical with the resonant or minimum impedance frequency.
The critical value at which resistance becomes too high for free oscillations to exist m a condenser-and-inductance cir- cuit, is almost never encountered in ra- dio transmitters. It may be computed from a simple rule, as follows: (1) Di- vide the total circuit inductance, in hen- rys, by the total capacity in farads, (2) take the square root of this ratio, and (3) multiply the result by 2. The re- sult is the "critical resistance" in ohms. For the antenna circuit of Fig. 1, this is found to be (1) 0.000373 henrv di- vided by 0.0000000012 farad = 3 10^000 ; (2) the square root of this is 556; (3) 2 times 556=1112 ohms. Thus if the
��resistance is less than 1112 ohms, the result of the condenser discharge will be oscillations at the rate of 236,- 000 per second; of course no ordinary sending circuit ever reaches so high a resistance value, so oscillations are al- ways to be expected. In receivers, how- ever, when detectors may be placed di- rectly in series within the circuit, the direct-current resistance is often seve- ral thousand ohms. Free oscillations cannot exist in such circuits, but a defi- nite tuning effect for forced oscillations is present, since, by adjusting the ca- pacity and inductance reactances to neu- tralize, the greatest alternating current can be made to flow.
Referring to Fig. 2, it is obvious that for a given charge in the condenser, the greatest current will flow when the re- sistance R is of the smallest value. It is also true that the oscillations will per- sist for the longest time when this re- sistance is smallest. The actual resist- ance in circuit may be made only that of the wires and spark-gap, so that the free oscillations may be made to vibrate back and forth hundreds of times for each spark. In an antenna like Fig. 1, however, the effective resistance can- not be reduced indefinitely, because in addition to the spark-gap and wires forming the inductance and leads, the ra- diation of energy in electromagnetic waves adds a few more ohms. Because of this, and also because the capacity of an antenna cannot be increased indefi- nitely without great expense, the two circuits of Figs. 1 and 2, are often com- bined in the arrangement of Fig. 3. Here the coil in the closed circuit, Lj, forms the primary of a transformer whose secondary is coil L2 in the open or antenna circuit.
When condenser C is charged and al- lowed to discharge through the closed circuit, free oscillations are produced of the frequency determined by the effec- tive capacity and inductance of the cir- cuit. In passing through the primary Lj, these free oscillations induce alter- nating voltages of their own frequency in the secondary coil L2 and the con- nected antenna circuit A Li L2 E. By adjusting the inductance of the secon- dary and loading-coils, so as to neutral- lize the capacity reactance of the an-
�� �
