Page:Popular Science Monthly Volume 88.djvu/327

 Popular Science Monthly

��299

��and the condensers D6 so that the sound of the dots is somewhat different from that of the dashes. This gives in effect a two-tone system, and obviously per- mits higher signaHng speeds than does the usual dot-and-dash method. In ad- dition to the increase in speed the two- wave lengths feature oft'ers excellent se- curity from interception of the messages by ordinary radio receiving stations.

When the signals received are sutti- ciently strong to operate a sensitive relay it is possible by this method to make a siphon re- corder pen-and-ink record correspond- ing exactly to cable "slip." If a relay is connected to each side of the receiv- ing system, the two contacts may be used to control a third polarized re- lay which will re- main in an open neutral position so long as no signals are received, but which, when waves are ar- riving, will close its local circuit and per- mit current to flow in one direction or the other according to whether it is operated by a dot-impulse or a dash- impulse. A siphon recorder in this last- named local circuit will recoid the sig- nals by a wavy line having a hump above its neutral position along a central line for each dot, and a hump below for each dash. Fig. 3 shows the actual connec- tions of apparatus set up to accomplish this result, and in this diagram the action may easily be traced from the sensitive relays U U, which are connected to the two detectors, to the siphon recorder g.

U. S. Patent No. 1127921, issued to G. W. Pickard, is on an important detail of receiving tuning apparatus. Before the adoption of inductance varying ar- rangements similar to that shown in this specification it had been customary to rely upon either sliding contacts, vario- meters or roller inductances for tuning. Each of these methods has disadvan- tages; sliders give poor contact at times.

���Fig. 3. Relay connections for recorder operation

��and cause loss of energy through short- circuited turns; variometers are limited in range of adjustment, and have their total resistance in circuit even at mini- mum inductance ; roller arrangements are bulky, and slow in operation. All these difficulties may be overcome by the use of multiple-point switches connected to the turns of the coils, but it would be practically impossible to have a switch- point for each sin- gle turn of a long coil. If a saving in the size of the switch is attempt- ed, by making each point cover a num- ber of turns, it is not found possible to get sharp enough tuning unless a n auxiliary variable inductance or con- denser is used.

The plan of wiring shown in Fig. 4 makes it possible to get sin- gle-turn steps of inductance on a long coil by using two small switches. One of these, indicated l)y S, has taps taken off the body of the coil at each tenth turn. The other. Si, has its points connected to each of the last ten single turns on the coil. The leads to the coil, A and G, run to the levers of the two switches ; and each ter- minal may be connected directly to the tenth turn of the coil by placing its re- spective switch lever on a button marked "O." This common zero of the two switches seems to be the novel point in

���Fig. 4.

��Common zero inductance coil

��switches on

��the present patent, and is the artifice by which it is possible to adjust to any in- ductance from zero to full value by steps

�� �