The syphon or siphon recorder is an antique electromechanical device used in telegraph offices invented by William Thomson, 1st Baron Kelvin in 1858. It was used to automatically record an incoming telegraph message, as a wiggling ink line on a roll of paper tape. Later a trained telegrapher would read the tape, translating the pulses representing the "dots" and "dashes" of the Morse code to characters of the text message.
In large telegraph networks the syphon recorder largely replaced the telegraph sounder, a device which produced "click" sounds representing the Morse code dots and dashes. With a recorder, a skilled telegraph operator did not have to constantly monitor the line waiting for messages to come in.
How it works
The siphon recorder works on the principle of a D'Arsonval galvanometer. See diagram. A light coil of wire is suspended between the poles of a permanent magnet so it can turn freely. The coil is attached via two wire linkages to the metal plate siphon support, which pivots on a horizontal suspension thread. From this plate a narrow glass siphon tube hangs down vertically with its end almost touching a paper tape. The paper tape is pulled by motorized rollers at a constant speed under the siphon pen. Ink is drawn up from a reservoir into the tube by siphon action and comes out a tiny orifice in the end of the siphon tube, drawing a line down the moving paper tape. In order not to affect the motion of the coil, the siphon tube itself never touches the paper, only the ink.
The current from the telegraph line is applied to the coil. The pulses of current representing the Morse code "dots" and "dashes" flowing through the coil create a magnetic field which interacts with the magnetic field of the magnet, creating a torque which causes the coil to rotate slightly about its vertical suspension axis. The wire linkages cause the siphon support plate to rotate about its horizontal axis, swinging the siphon tube across the paper tape. This draws a displacement in the ink line on the tape as long as the current is present in the coil. Thus the ink line on the tape forms a graph of the current in the telegraph line, with displacements representing the "dots" and "dashes" of the Morse code. An operator knowing Morse code later translates the line on the tape to characters of the text message, and types them onto a telegram form.
Kelvin's electrostatic syphon
The siphon and an ink reservoir are together supported by an ebonite bracket, separate from the rest of the instrument, and insulated from it. This separation permits the ink to be electrified to a high potential while the body of the instrument, including the paper and metal writing tablet, are grounded, and at low potential. The tendency of a charged body is to move from a place of higher to a place of lower potential, and consequently the ink tends to flow downwards to the writing tablet. The only avenue of escape for it is by the fine glass siphon, and through this it rushes accordingly and discharges itself upon the paper. The natural repulsion between its like-electrified particles causes the shower to issue in spray. As the paper moves over the pulleys a delicate hair line is marked, straight when the syphon is stationary, but curved when the siphon is pulled from side to side by the oscillations of the signal coil.
Power to pull the roll of paper tape through the syphon recorder was usually supplied by one Froment's mouse mill motors. These also drove an electrostatic machine to generate the electricity to power the syphon.
Muirhead's vibrating recorder
A simpler mechanism was developed by Alexander Muirhead. This used a vibrating pen to avoid the same problem of the ink sticking to the paper. The recording pen was suspended on a thin wire, vibrated by an electromagnet and break contact, similar to that of an electric bell.
- Byrn, Edward (1900). The Progress of Invention in the Nineteenth Century (1900 ed.). New York: Munn & Co. pp. 36–37.
- Kennedy, Rankin (1903). Syphon Recorders. Electrical Installations. Vol. V (1903 five-volume ed.). London: Caxton. pp. 79–81.