Shown here as it appeared 75 years ago is a view of 100 kW shortwave stations WGEA and WGEO, the General Electric station at Schenectady, New York.  The picture appears in the November 1941 of FM Magazine, along with an extensive look at the technical details of the stations, as well as those of GE’s sister station in San Francisco, KGEI (about which we wrote earlier).

The station had its start in 1925 with an experimental license under the call sign W2XAD. When the FCC deemed shortwave broadcasting to be sufficiently developed for commercial service in 1939, the station went on the air commercially.

GE received over 30,000 pieces of mail per year, and prepared its programs based on listener preferences. In addition to English, the stations had regular programs in Spanish, Portuguese, French, Dutch, Czech, and Chinese.

Frequencies used at the time were 6190, 9530, 9550, 9670, 15330, and 21590 kHz. Directional antennas beamed programs from Schenectady to Central America, Brazil, Europe, Argentina, and North Africa, and from San Francisco to South America and Asia.

One thing that sometimes mystifying when looking at diagrams of early spark transmitters is that the descriptions often lack much detail as to the most important component, the induction coil or spark coil. A good description can be found in Electrical Experimenter magazine, November, 1916, from which this illustration is taken.

The “coil” is really two coils, plus an interrupter, and the operation is described here:

When the primary switch of such a coil is closed, the battery current passes through the first winding on the core and magnetizes it. This attracts the iron armature on the vibrator spring, as shown in Fig. 1, and when this spring breaks contact with the platinum tipped screw in front of it, the circuit is opened. At this juncture there is induced in the secondary winding a very powerful current. The spring-actuated vibrator returns to its former position in the fraction of a second and the process is repeated all over again.

The referenced schematic diagram is shown here:

The plans for this simple two-tube set appeared eighty years ago this month in Radio Craft magazine, November, 1936.

With plug-in coils, the set tuned 10 to 600 meters, and the author reported picking up numerous European and other foreign stations in New York City with excellent volume. The set employed two tubes, either type 37 or 76. One served as rectifier, with the grid and plate terminals tied together. The other served as the regenerative detector which was “capable of picking up even the faintest of signals.” An antenna of 30-100 feet was recommended. No ground connection was necessary, since the house current took care of the ground.

The duo shown here are putting together the one-tube receiver described in the November 1941 issue of Popular Mechanics.

The set was billed as being for the beginner who wanted to graduate from the crystal set stage, or for the radio experimenter who enjoyed pulling in some DX, an ability this set possessed.  The single tube was a 117N7-GT, so it required no A battery.  Instead, the filament was powered by the AC line.  The set was so efficient that it would run on a 30 cent 4.5 volt battery supplying the B+ voltage.  The required parts, including the tube and battery, but not including the headphones, priced out at about $4.50, even accounting for recent price increases.

A hundred years ago, the October 1916 issue of Popular Science gave detailed instructions for constructing a transmitter capable of sending one mile, the diagram of which is shown above. The set was powered by nine volts, consisting of twelve dry cells wired in series-parallel.

When the modern reader looks at a diagram such as this one, it’s not readily apparent how the set could actually work, since one critical part of the transmitting coil is not shown. The article noted that plans for constructing the coil were available in many books, but recommended that it was simpler and cheaper to simply buy one. It suggested an automotive spark coil, which could be obtained from a garage or electrician at nominal price. The article noted that even a new coil would cost no more than three or four dollars.

Such a coil is actually a transformer, and the primary is wired at the factory in series with an normally-closed interrupter. When the coil is energized by the nine volts, it creates a magnetic field, which opens the interrupter contact. This breaks the circuit, and the coil de-energizes, which results in the interrupter closing again, allowing the process to repeat. The net result is that an interrupted direct current is produced in the primary, and this induces an alternating current in the secondary. The secondary is a large enough voltage to make a one-inch spark in the attached spark gap.

The spark gap, shown here, is constructed from two pieces of zinc mounted on a hardwood base boiled in paraffin.  The key could be an ordinary telegraph key, which sold for about 75 cents, or could be homemade.

The article stressed that the loading coil, shown here, was required to keep the transmitter on the correct wavelength, below 200 meters, to comply with federal law.  It was constructed between two square boards, about 12 by 12 inches.  FInally, an aerial of no more than 75 feet was required, and the article contained pointers for its installation as well.

Finally, the article described operation of the set:

Whenever the key is pressed, if the set is properly connected and adjusted, a bright, snappy, singing spark will jump across the gap. Each spark starts a train of high frequency currents oscillating back and forth in the aerial wires, and a train of electromagnetic waves is radiated into space. A suitable wireless receiver located where a portion of these radiated waves will reach it, will pick up some of their energy and produce from it a sound which indicates the dot-and-dash buzzes of a Morse signal.