by John S. (Jack) Belrose, VE2CV, VY9CRC
Peter Bouliane, VE3KLO

b-logo.gif (1352 bytes)
Collins Collection
The Museum
War Years
Early years
Hamond History
Visit Us

In 1906 Reginald Aubrey Fessenden, using an HF Alternator, modulated by an asbestos covered microphone in series with his tuned antenna system, made a pre-announced (3-days earlier) Christmas Eve broadcast, to US Navy Warships, and ships of the United fruit Company, using Fessenden type receivers

The Christmas Eve program as recounted by Fessenden consisted of:

"... first a short speech by me saying what we were going to do, then some phonograph music.... Then came a violin solo by me ... which I sang one verse of, in addition to playing the violin, though the singing, of course, was not very good. Then came the Bible text, Glory to God in the highest and on earth peace to men of good will, and we finally wound up by wishing them a Merry Christmas and then saying that we proposed to broadcast again on New Year's Eve. "

We have created a simulation of that broadcast, using a tabletop electromechanical HF alternator, operating at a frequency of 48 kHz. The carbon microphone is a cartridge from an old Bell telephone set. The electromechanical generator comprises a DC motor driving a 2 x 400 step/revolution stepper motor. This is actually an 800 step/revolution stepper motor, salvaged from a scrap hard disk drive. When driven as a generator, it produces a reasonable sine wave --- each winding produces a signal in phase quadrature with the other --- generating a frequency 200 Hz for each revolution. So as an HF alternator, it can produce quite high frequencies when rotated at rather high speeds. Driven at 14,400 rpm by a DC motor, it generates a frequency of 48 kHz. This HF alternator is connected to the rest of the circuit by a long coaxial cable, because of the mechanical noise it produces!

The idea of simulating an HF alternator, like the type used by Fessenden, came from an article and discussions with J.R. (Jim), Moritz, MBMU, who used similar hardware to devise an electromechanical receiver for VLF [see On-Line http://wireless.org.uk/mechrx.htm] . In fact Jim sent us two 2 x 400 step/revolution stepper motors for our project.

A photograph of our HF alternator is shown in the figure below (size of base plate 5.4 x 8.8 inches). The step-motor is the component on the left. For our experiment we used just one of the windings of the stepper, which we terminated in 50 ohms. The unbalanced-to balanced ferrite toroidal transformer is coupled by a 50-ohm coaxial cable to an equivalent circuit of the antenna system used by Fessenden (L-C-R not unlike the umbrella top loaded 420 foot tower antenna system used). The carbon microphone is in series with our equivalent antenna circuit. The signal across the resistive component is coupled to an untuned receiver circuit, and recorded on a Macintosh PowerBook G4.

The DC motor we used was never designed to operate at such a high speed, and the stepper motor was not designed to be driven at such a high speed. We should have had a brushless motor, since the motor we have (during the period of time used) is starting to exhibit momentary interruptions, due to the wear on the brushes.

The audio clip for our simulated broadcast starts out with the voice of author Belrose, which is then followed by clips taken from the CBC-Shell Oil Company 1979 drama on Fessenden. This drama was one of a series entitled The Winners, and the particular drama was entitled "The Forgotten Genius". The HF alternator was in one room, the microphone, the circuitry simulating the antenna, and the receiver were in another room --- in spite of this one can hear on the recording the subdued high pitch whine of the electromechanical HF alternator. We decided to record the file this way --- so the listener can perhaps judge that the recording is genuine. There is one brief dropout (during the period when Fessenden is playing his violin) where the brushes momentarily skipped.  Click the Play button below to listen.

What have we done?
Clearly the quality of the sound for our simulation of the broadcast is perhaps too good. Why?

We have by our hands-on experience shown that undamped LF waves (continuous waves) can indeed be generated by a machine, but not without considerable difficulty; and indeed a carbon microphone as used by Fessenden does indeed modulate the signal rather well. But we have not simulated the effect of the narrow bandwidth of an efficient electrically short antenna system. The estimated bandwidth of Fessenden's antenna (a 420 foot umbrella loaded tower), at 80 kHz, the frequency he probably used, is about 1 kHz, which is far two narrow for AM broadcasting. But recall that a carbon microphone is inserted in series with antenna impedance, and this introduces a dominant loss resistance and hence increased bandwidth. The "dynamic" operational bandwidth is probably four times the antenna system bandwidth (and the radiation efficiency is reduced by a similar factor) --- "dynamic" because the bandwidth varies with the level of modulation. The bandwidth of antenna systems used with present day long-wave transmitters, while a problem of concern to broadcasters, is constant, since the signal feeding the antenna is modulated.

The bandwidth of our experimental system is large compared with the bandwidth for the real antenna, since the Q-factor of the small inductors we used is certainly low compared with the Q-factor of the large helical coils used to tune real world long wave antenna systems. Whatever, we enjoyed doing this project. We certainly experienced problems in doing what Fessenden did. Fessenden did not have a problem with the rotational speed of his AC motor, since he used a gearbox to step up the shaft speed of the HF alternator. But certainly, since his HF alternator had to operate at similar speeds as for our experiment, the shafts have to be exactly aligned (we used a precision small flexible coupling), and the bearings have to be well lubricated (Fessenden used an oil pump), and there has to be no wobble or shake with shaft rotation. And, Fessenden had an additional problem, keeping his carbon microphone cool, and not fusing the carbon granules, since the transmitter powers used were orders of magnitude greater than our tiny experimental model.

John S. (Jack) Belrose, VE2CV, VY9CRC
Peter Bouliane, VE3KLO
Communications Research Centre Canad
Ottawa ON K2H 8S2

Other Reference Info at http://www.radiocom.net/Fessenden/

  [ Top ]

Last updated July 02, 2008

1998 - 2004 Hammond Museum of Radio

Web space courtesy of Information Gateway Services