Electrical Experimenter, April, 1920, pages 1281, 1326:

Some  Long  Distance  Radio  Telephone  Tests

A SERIES of radio telephone tests and experiments by the writer performed during the months of December and January have produced some results which may be interesting to the readers of the ELECTRICAL EXPERIMENTER. The results are interesting, not only because of the records made, but of the peculiar conditions under which the experiments were performed. Robert F. Gowen
    The transmitting equipment consisted of a standard one kilowatt deForest Oscillion panel transmitter which is designed to work on large ship type aerials on wave-lengths between 500 and 1000 meters. This transmitter uses two ½ kilowatt power tubes for the generation of the high frequency currents. The plate current of these tubes is supplied from a motor-generator set which delivers a voltage of 1500. The tubes are of a new type made of Pyrex glass and the plates are of Tungsten so that external cooling is not required. The tuning is done by means of a variable sectional series condenser and by varying contact clips on a helix in the rear of the panel. The circuit employed was of the standard deForest single tuned type with capacity coupling between the grid and plate circuits.
    A special new type of modulating circuit developed by Mr. C. V. Logwood was installed, the novelty being in that both the grid and plate circuits of both tubes are modulated by the voice currents at the same time. Continuous wave telegraphy is accomplished by making and breaking the grid leak circuit and a modulated wave was added by substituting with a switch, a buzzer and key in place of the microphone.
    The transmitter was installed at the home of the writer in Ossining, N. Y., and connected to a comparatively small aerial, four wires 170 ft. in length with a spacing of 30 inches and with an average height of 55 feet. A 60-foot lead-in was tapped off the end nearest the station and brought to the transmitter through a hole in a window pane. This aerial has a capacitance of approximately .0007 micro-farad and a fundamental wave-length of about 355 meters.
    The ground system consisted of 84 square feet of copper sheet buried beneath the foundations of the operating room about 8 feet below the transmitting apparatus. There were also additional grounds to the water pipes, gas pipes, heating apparatus, etc., of the house.
    Upon installation, no difficulty was experienced in making the transmitter operate successfully on 750 meters for tests and demonstrations with the deForest laboratories at High Bridge, but when later, experiments were made to cause the set to oscillate at 200 meters in order to work with the amateurs, it was found that the constants of the circuit were such as to make it decidedly unstable at this high frequency. After a good deal of experimentation and adjustment, however, fairly good results were obtained at a wave-length of 330 meters which, it will be noted, was below the fundamental of the aerial.
    Using this wave-length during December the writer attempted to communicate with amateurs in the vicinity of New York and found great difficulty in making the telegraph signals heard because of the extremely sharp tuning of the emitted wave and also because very few amateurs listen on wave-lengths over 250 meters.
    On the 330 meter wave-length the transmitter would put but 2.4 amperes into the aerial with an input of 300 watts, tho a maximum of 2.6 amperes could be obtained with an input of approximately kilowatt. Owing to the unstable conditions caused by the circuit constants the oscillator had very poor efficiency and about January 1 the writer was surprised to have a report that the signals were being heard at Utica about 175 miles away. Mr. Benas at Utica reported by radio telegraph and by letter that the music sent out at the Ossining station could be heard 100 feet away from his telephone receivers when using a two step amplifier, so that in two instances, his friends there had used it as a dance orchestra. Robert K. Trump
    It was apparent from this report that altho it was difficult to get amateurs to listen and answer the calls on 330 meters, when the signals were once tuned in, they were very loud. This incentive prompted the idea of working later at night when the amateur interference was less, and shortly thereafter Radio Station 8ER answered a call and reported that they were at St. Mary's Ohio, more than 600 miles away.
    Continued efforts in adjustment of the circuit and methods of calling then brought in reports from many stations thruout Ohio, and Indiana. Reports of successful reception of the speech were received also from Wakefield, Massachusetts, Gaffney, South Carolina, Napanee, Ontario, about this time, increasing the distance to about 750 miles. Subsequently a letter was received from the operator on a yacht 300 miles south of Jacksonville, Florida reporting good reception when using only one receiving tube.
    While continuing tests with Mr. Charles Candler at St. Mary's, Ohio, the signals and speech were picked up by Mr. R. H. G. Matthews, Chief Engineer of the Chicago Radio Laboratory at his Chicago Station. He reported that the speech could be heard plainly at a distance of over 10 feet from the telephone, altho the buzzer on the modulated wave was slightly louder. The power input for these tests was 285 watts.
    A continuance of tests with Mr. Matthews during the last week in January, brought in reports from Little Rock, Arkansas, and from Valley City, North Dakota, distances of about 1200 and 1500 miles respectively. Mr. Clayton at Little Rock reported the speech particularly understandable, but Mr. Pray at Valley City, North Dakota, stated that the speech was weak, but that he was using a very inefficient receiving set. In February, the speech was successfully heard and copied at Baudette, Minnesota, which is nearly as far away as Valley City, and many letters have been received from St. Paul, St. Louis, Minneapolis, and many cities in Michigan to the same effect. In the latter instances, however, the input was increased to 510 watts and in one or two cases the tests were made on a prearranged schedule.
    In practically every instance the reports were most enthusiastic as to the modulation or clearness of the speech. Whole spoken sentences were quoted and some even went so far as to comment upon the accent of the person talking. The reports unanimously indicated that once the receiving station found the tune of the transmitted wave, there was no difficulty in hearing the speech clearly and loudly. Mr. Candler at St. Mary's reported one night that he could not use his two steps of amplification as it hurt his ears. He telegraphed back that the writer's voice fairly screamed at him.
    The experiments have shown that with a remarkably small power input, stations practically half way across the United States can be reached even when the set is operating very inefficiently on an aerial and on wave-lengths for which it is not designed.
    The tests are being continued, at eleven o'clock nightly and the writer would be pleased to hear from those who are interested as to their results in receiving him at some distance.
* Engineer deForest Radio Tel., and Tel Co.