I have occasionally visited radio stations that seemed particularly interesting. On most of my many voyages on ships, I have got to know some of the Radio Officers, and of course, to visit the ship's radio room. On the Cunard ship Samaria, crossing from New York to Liverpool, I was carrying the large home brewed superhet referred to under "Short Wave Listening". The ship's company tied a wire to a mast stay and threaded it down through the ship to my stateroom so that I could use it as an antenna for my radio! It worked quite well.
St. Andrews, where I graduated from University, is in eastern Scotland only a few miles from a village called Cupar. From the mid 1920's on, Cupar was the British receiving station for the first transatlantic telephone. This little known circuit was on long waves, 5000 meters (60 kilohertz). The eastbound signals went from Rocky Point, Long Island to Cupar, and the westbound went from a station at Rugby, England to a receiving station in Maine. The Cupar receiver used a grid of around a hundred center fed antennas, each a couple of hundred feet long stretched between tall wooden telephone poles. The whole array occupied about a mile square. The signals from this network of antennas were taken on transmission lines to the receiver. Phasing circuits made the array quite directional towards New York. The receiver looked more like a telephone central office than a radio receiver.
I visited Cupar in the mid 1930's. Years later I obtained permission from RCA Communications, the subsidiary of the Radio Corporation of America which dealt with point-to-point communication, to visit their transmitting station at Rocky Point. The station contained several dozen short wave transmitters, each of 10 to 20 kilowatts power. One of them, still in service after 20 years or so, was the first one purchased by RCA from Marconi shortly after Marconi discovered that short waves from transmitters of a few thousand watts could span distances requiring hundreds of thousands of watts on longer wavelengths. Most of the Rocky Point short wave sets transmitted international morse telegraph messages to points all over the world, though some were used for Teletype service, and a few for voice or music. The latter relayed news and such for broadcasting abroad. The corresponding west coast station at Bolinas, California, relayed network radio shows like Jack Benny, Fred Allen and the Lucky Strike Hit Parade to Hawaii (if I missed the show on Eastern Time I could catch it on its repeat on Pacific time on the short wave relay). At that time undersea cable transmission of voice or music was impossible.
At Rocky Point I also saw the long wave radio telephone transmitter which was aimed at Cupar. A.T.& T. rented space for the transmitter and an antenna from RCA. It was a single sideband suppressed carrier transmitter capable of 250,000 watts output. As I remember, the upper sideband was used for the eastbound circuit and the lower sideband for the westbound.The transmitter used twenty big water cooled tubes in its final amplifier, and the tuned circuits were room sized. I asked the operator what happened if one of the tubes failed. He said "Nothing. Nineteen is just as good as twenty!" Actually these tubes lasted for years and years. The transmitting antenna was made up of a number of parallel cables nearly ten miles long, supported on the tops of steel towers 700 feet high. A few years after the long wave transatlantic telephone circuits were established, the telephone service took up short wave. Until successful telephone cables were developed for transoceanic use, the long wave equipment was kept in working order for the occasions when solar storms interrupted short wave long distance communication. At these times, the long wave signals usually got stronger!
I also visited an RCA long wave radiotelegraph station at New Brunswick, NJ. It used a huge antenna hundreds of feet high and miles long like the one for the long wave telephone at Rocky Point. The transmitter was an Alexanderson alternator, the invention of a General Electric engineer, with an output of 400,000 watts. It looked like an ordinary power station generator. Its frequency was around 17 kilohertz, which made its wavelength around 17,500 meters or 11 miles. This station, whose callsign was WTT, had a thirty sixth harmonic (or overtone) which could be heard sometimes interfering with the broadcasting station WEAF in New York (660 Khz) when I listened to it in Philadelphia. A similar transmitter of nearly a million watts was built in the Netherlands to try to communicate with the Dutch East Indies (now Indonesia) directly by long waves just as short wave was being developed to do the same thing with a fraction of the power, but without the extreme reliability of the long waves.
Commercial radio communication of this kind, either on short or long waves, has almost disappeared. First telegraph cable speeds were increased. Then telephone signals were sent by cable when tiny long-lived vacuum tubes were developed that could be sunk beneath the ocean to amplify them. The transistor made the telephone cables amplifiers cheaper and more reliable. Satellites launched into orbit around the earth allowed not only telephony but television to move from continent to continent. Now the satellites are challenged for many purposes by fibre optic cables. Ships are using satellites, so that cargo ships no longer carry radio officers. For a few more years short wave radio will continue to be used to talk to transoceanic airliners, but soon satellites will fill this need too.