��Popular Science Monthly
��through the wire, and there is always difficulty in maintaining good contact. Further the slider usually short circuits several turns at its point of contact, and thus causes mistuning and loss of signal strength. These disadvantages of the slid-
���The rotary switches mounted on the panel, one having ten and the other fifteen points
ing contact, taken together with the poor circuit design often employed, have brought the direct-coupled circuit of the sort described into disrepute; there is really little choice between the direct and in- ductively coupled receivers, however, pro- vided that both are properly built.
By making the loading and transformer coils in separate units, and fitting each with switch contacts instead of sliders, it is entirely practicable to produce a receiving installation which has all the selectivity of the inductively coupled type and still avoid some of its disadvantages. Such coils are described below, and their mode of connection is shown in Fig. 7, which cor- responds to Fig. 5 of the December article.
The Loading Coil This instrument is shown in Fig. i. The basic piece is a paper tube of about 4 in. diameter and 8 in. length. Beginning about I in. from one end. No. 20 gage double silk-covered magnet-wire is wound on evenly for about 6 in., which will take 149 turns. A tap must be taken out for each of the first ten turns, and one for each tenth turn thereafter, as shown in Fig. 4. The best plan is to fasten the end of the wire, before beginning winding, through two small holes punched in the paper tube, leaving about 2 in. of wire free for the tap marked "9" in Fig. 4. Then a single turn is wound on, and a small loop twisted in as shown in Fig. 2. This twist and loop stand up from the surface of the coil, and the wire leading to contact "8," Fig. 4, is later soldered to the loop. The second turn of wire is then wound on the
��tube, and another twisted loop for the contact "7" is made. Thus a twist is put in for each of the contacts, at the end of each turn, till that marked "O" is reached. Then ten turns are wound without a tap, the twist for "10" being taken out at the 19th turn of the whole coil. Similarly taps for "20", "30", etc., are made at each tenth turn thereafter.
The completed coil is to be mounted between two end blocks of wood or hard rubber. A, in Fig. i. Three small pieces are fastened to the inside faces of each of the end blocks, as shown in dotted lines, to keep the coil from slipping sidewise; and the whole is held together by a piece i in. square passing through from end to end in the center of the coil. Screws B, with washers under their heads, pass through holes bored in the end blocks and clamp the tube by threading into the ends of this central stick.
As further shown in Fig. i, a switch- panel is mounted on top of the end blocks. This should be made of hard rubber or fiber, about 3^ in. thick, though hardwood /^ or 3^ in. thick will do. It is fastened to the end blocks by means of two wood-screw binding posts D, D^, Fig. i and 3, and the screws E, Fig. 3. On it are mounted two rotary switches, one having 10 and the other 15 points, as shown in Fig. 3. Any type of switch-arm will do for these, but the easiest manipulation will be obtained if a center-knob type is used.
The ID-point switch should be markec "Units" and the 15-point "Tens"; the
���The wiring diagram for the loading coil showing the turns and where they are tapped
buttons are to be numbered from zero to 9 for the former and from zero to 140 in tens for the latter. The taps from the coil itself are to be connected with these switch- points by means of short soldered leads, as shown in Fig. 4. Great care must be taken to see that no short circuits are made as this wiring is put in place. The central points of the switches, i. e., their arms, have wires leading to the binding posts