Page:Popular Science Monthly Volume 88.djvu/646

 618

��Popular Science Monthly

��er or one of the telephone leads. Some correspondents have suggested perma- nently connecting one of the telephone tips to the telephone diaphragm by use of a small metal strip, saying that by trying the several possible combinations, they were successful in finding an arrangement which actually reduced the interference without weakening the sig- nals. A variation of this method is to use the metal-capped telephones and to ground the cases, either directly or through a condenser. Sometimes it is sufficient to connect the case or one of the telephone terminals to the operator's body by bending a piece of tinfoil over the ear-cap and connecting it to the case or cord tip. It has also been found use- ful to connect together the metal cases and headband, as well as to connect to- gether a certain one of the four tips and the metal case. Which cord-tip to connect in this way must be ascertained by trial.

It has been learned that at a number of stations both the induction and the arc noises can be reduced if the antenna is changed so as to point directly away from the power lines. When the aerial and the no- volt wires are parallel there is, of course, the greatest inductive effect between them, and when they are about perpendicular the induction is least. Sometimes a position not exactly perpendicular gives the smallest inter- ference, because of an irregularity in the magnetic field around the power wires.

Taken as a whole, the elimination of arc noises remains a big problem at some stations. In many cases the remedies suggested above will reduce the disturbances so much that they will cause no trouble, but it is likely that at some other stations the interference will persist in spite of the hardest work to get rid of it. If it becomes necessary, the "balanced primary" method may be tried as a last resort; in this arrange- ment a small extra antenna is erected near the power lines and connected to ground through a second primary which opposes the effects (on the secondary) of the regular primary coil connected to the regular receiving aerial. This more complicated circuit may prove worth while, since, by its use, the noises have, in some cases, been almost entirely cut out after the simpler plans failed.

��A Crystal Detector

THE main advantage of this detector is that a great number of sensitive spots on the crystal are obtained. A co- herer-stand may be used, by substituting

��Crystals

���r&it-

��A great number of sensitive points on the crystal are obtained

small galena crystals for the filings. In the absence of a coherer-stand, use two large binding posts, two brass rods and a piece of glass tubing which will fit snugly over the brass rods. The mineral is prepared by placing a sensitive piece of galena in a small piece of cloth and then pounding the crystal into small pieces. The small crystals are then sep- arated from the powdered mineral, and placed in the glass tube. Adjust the de- tector by turning the glass tube and mov- ing the brass rods until the maximum sensitiveness is obtained. A buzzer is used to excite the circuit in order to se- cure the best adjustment easily. Great care should be taken not to touch the crystals with the hands, since moisture or grease will decrease their sensitive- ness to a great extent.

Restoring Electric Light Bulbs

OFTEN when the electric lights of the tungsten filament type go out, or burn out, it is caused by the break- ing of the filament wire. When this is the cause, screw the bulb into the socket of a flexible cord and turn on the cur- rent. By holding the bulb in a hori- zontal position, manipulate it by turning and rolling and tapping it with the hand to cause the filament wires to cross. When successful it will instantly light. While lit, hold it in a quiet position for a few minutes till the wires weld, after which it can be used for regular service. The writer has been able to recover over 60% of light bulbs he has tried, and they have lasted from three hours to three weeks. — John Hoeck.

�� �