The Music of Radio is a history series showcasing the relationships between radio and electronic music. This installment focuses on sounds created by arcs in the days before incandescent lighting cast its long and overshadowing glow.
The first source of electrical lighting was the arc lamp. It was also used as a means for producing an electrical form of singing. Invented by Humphry Davy in the first decade of the 19th century the arc lamp created light from the electricity passing between two carbon electrodes in free air. To ignite a carbon lamp the rods were touched together allowing a low voltage to strike the arc. They were then drawn apart to allow the electric current to flow between the gap. This first means of electrical lighting also became the first commercial use for electricity beginning around 1850 but it didn’t really take off until the 1870’s when regular supplies of power became available.
Three major advances in the technology occurred during the 1880’s that helped to spread the adoption of the arc lamp. The first was a mechanism to automatically adjust the electrodes. The second was the placement of the arcs in an enclosure to cause the carbon to burn at a slower rate. Last salts and tiny amounts of metals were added to the carbon to create flames of greater intensity and different colors. At this time a number of companies became involved with the manufacturing of these lamps and they began to be used for lighting on streets and other public places. Yet there was one feature about the light source that many folks found disagreeable. These were audible power-frequency harmonics caused by the arcs negative resistance. Nikola Tesla was one of the guys who set to work on this problem of noise. In 1891 he received a patent for an alternator that ran at 10,000 cycle per second that was to be used to suppress the undesirable sounds of humming, hissing and howling emitted by the lamp.
Tesla’s invention must have been impractical or just never caught on because over in London in 1899 the Victorian electrical engineer William Duddell had been appointed to tackle the problem of the lamps dissonant electrical noise. Duddell was an illuminated man and he took a different angle than Nikola. Instead of suppressing the sounds he transformed them into music. In the course of his experimentation Duddell found that by varying the voltage supplied to the lamps he could control the audible frequencies by connecting a tuned circuit that consisted of an inductor and capacitor across the arc. The negative resistance of the arc was excited by the audio frequency oscillations from the tuned circuit at its resonant frequency. This could be heard as a musical tone. Duddell used another one of his inventions, the oscillograph, to analyze the particular conditions necessary for producing the oscillations. He demonstrated his invention before the London Institution of Electrical Engineers by wiring up a keyboard to make different tones from the arc. Being a patriotic fellow he played a rendition of God Save the Queen. His device came to be known as “Singing Arc” and was one of the first electronic oscillators. It was noted that arc lamps on the same circuit in other buildings could also be made to sing. The engineers speculated that music could be delivered over the lighting network, but this never became a reality. Duddell toured his instrument around Britain for a time but his invention was never capitalized on and so remained only a novelty.
Duddell’s Singing Arc had been very close to becoming a radio. Marconi’s spark-gap transmitter had already been demonstrated in 1895, yet Duddell thought it was impossible to leverage his Singing Arc to produce radio frequencies instead of audio frequencies. The AC current in the condenser was smaller than the supplied DC current so the arc never extinguished during an output cycle, making it impractical to use as an RF transmitter. With this set up it was not possible to reach the high frequencies required for transmission of Radio-telegraphy. If he had managed to increase the frequency range and attached an antenna his invention could have become a CW transmitter.
Enter Danish physicist Valdemar Poulsen. In 1900 at the Paris World Fair this man demonstrated his inventive flair with the world’s first magnetic recording device, the Telegraphone. Working with another Dane, P.O. Pederson, he used his skills to improve upon Duddell’s Singing Arc. They were able to raise the efficiency of the device up to 200 kilohertz. His method of oscillation made use of an AC current from the condenser that was large enough to extinguish the arc but not so great that it caused the arc to restart in the opposite direction. (A third method of oscillation was used in spark gap transmitters where the arc is extinguished but might reignite when the condenser reversed.)
In 1903 they patented the Poulsen arc wireless transmitter that was the first generate to continuous waves, and one of the first pieces of technology to transmit through amplitude modulation. Poulsen’s version was used for radio work around the world up into the 1920’s when it became replaced by vacuum tube transmitters.
The method of operating a Poulsen arc transmitter required frequency shift keying. Straight on-off keying could not be used because of the time it took for the arc to strike and re-stabilize. With the arc staying on throughout operation the keying frequency needed to be adjusted anywhere from one to five percent. The signal at the unwanted frequency was deemed a compensation wave. Two keys were used, a “mark” or closed key, and a “space” or open key. This mode took up quite a chunk of bandwidth, as it also transmitted on the harmonics of the frequencies. Since around 1921 the use of the compensation wave method for CW has been prohibited. One way amateurs worked around this was to use a dummy antenna, or back shunt, tuned to the same frequency as the transmitter to absorb the load from the arc while keeping it running.
It as common to use audio frequency shift keying in telephony. Most early Bell modems used FSK at around 1200 baud. FSK is still used for some digital ham modes such as AMTOR, PACTOR, and GTOR.
P.S. For those interested in creating a high-voltage Plasma Arc Speaker based on Duddell’s Singing Arc an article on how to do just that was written for Make Magazine by John Iovine. The core of his project is a 555 timer and an insulated gate bi-polar transistor. Schematics, instructions and a video of it in operation are available at: http://makezine.com/projects/plasma-arc-speaker/.