Before vacuum tube transmitters

We saw earlier that the first Hertzian waves that could be produced, those generated by the oscillating discharge of a capacitor, had the property of damping very rapidly, their amplitude decreasing progressively and eventually becoming zero. For a long time, Wireless Telegraphy apparatus was based on this principle, and until around 1917, almost all oscillation generators were damped wave transmitters. However, when it was desired to perfect the technique of radiotelegraphy, particularly to increase the power of transmitting stations and multiply the number of radiocommunications, when it was desired to solve the problem of radiotelephony, significant difficulties were encountered with damped waves, and this led to the search for means of producing waves whose amplitude remains constant, and which were called "continuous" waves. This was achieved first with arcs; but the waves generated by these devices were not pure and caused troublesome interference. Continuous waves were then successfully produced using alternating current machines or high-frequency alternators. The waves obtained by this process are excellently suited for long-distance communications, but since the rotational speed of these machines is necessarily limited, it only allows obtaining waves of several thousand meters. The most modern method of generating continuous waves, which has superseded all others in most cases, is that which uses three-electrode tubes, which we will discuss later.

Before examining how Hertzian waves make it possible to transmit speech and music, it is necessary to recall in a few words the principle of ordinary wire telephony. It is known that the sensation of sound results from the action of air vibrations on our ear. Sounds are distinguished from each other by three essential characteristics: intensity, pitch, and timbre. The intensity of a sound depends on the amplitude of the vibrations. The greater this amplitude, the louder the sound. The pitch of a sound is determined by the frequency, or number of vibrations per second; the higher this frequency, the higher the pitch. Let us note here, in passing, that the sensitivity of our ear is limited in this respect. Only sounds whose frequency is between 12 vibrations per second and 12,000 are audible to us. Timbre is the characteristic that allows us to distinguish two sounds of the same pitch emitted

In the state of the art in 1919, only long waves of the order of several thousand meters could be considered for long-distance radiocommunications. Shorter waves, and in particular short waves, had, in fact, been used from the first Wireless Telegraphy experiments. The small devices then available produced them naturally. But as soon as it was desired, in order to increase the ranges, to increase the power of the transmitters, it became necessary to use increasingly longer waves to which the devices available were much better adapted. Moreover, it had been observed that short waves behaved very irregularly and were the site of very particular phenomena whose laws had not yet been elucidated and which compromised the security of long-distance links. The propagation of long waves was much better known, and it was known that it could take place fairly regularly day and night. To establish good permanent communication between two points, the difficulty consisted mainly in employing, with acceptable efficiency, high power. Thus, engineers sought the solution to the problem of long-distance radiocommunications in the realization of very powerful transmitting stations with very long waves. The principle of the Poulsen arc was first used. The Société Française Radioélectrique, for its part, had established a good implementation. But the arc offered a serious disadvantage inherent in its very nature, that of producing very numerous and very intense harmonics. The efficiency of this process was therefore mediocre; the power employed was effectively used only for a small part on the working wave, the rest being wasted in secondary radiations which, moreover, caused interference to other links.

The realization of this project then raised considerable mechanical difficulties: to obtain directly high frequencies, which had to reach at least 500 times those of industrial alternators, it was necessary to adopt unusual peripheral speeds, very small pole pitches, and extremely reduced air gaps. Now, too high a speed increases friction losses and causes abnormal stresses in the rotating parts, thus significant elastic deformations and dangerous vibrations that risk causing the detachment of the windings. If the air gap is very small, the precision of the construction must be extraordinary so that the rotor does not come into contact with the stator. The persevering and combined efforts of the Société Française Radioélectrique and the Société Alsacienne de Constructions Mécaniques de Belfort resolved this difficult problem with unparalleled success. Unfortunately, the war having interrupted this type of work, it was only in 1915 that the first machine (with a power of 5 kilowatts) was built and put into service at the Lyon (La Doua) station. At the end of 1918, a 125-kilowatt alternator was installed at the Lyon station (1).

one of the 500 KW HF sets of the Sainte-Assise station. In the center, the HF alternator. At the ends, the drive motors.

Below: the rotor of a 500 kW alternator. Massive steel part, capable of withstanding a traction of $$ kgs per square millimeter. It carries on its periphery thin sheets (thickness 0.03 mm) of mild steel, carefully enamelled, resisting a traction of 32 kgs per square millimeter. The safety coefficients that have been adopted for this essential part of the machine, whose peripheral speed reaches 150 meters per second, are extremely high. The French high-frequency and high-power alternator was created, and its qualities exceeded the most optimistic hopes (1). Its operation proved impeccable, the emitted wave was very pure, completely devoid of harmonics and very stable. This type of alternator has been built for various powers, from 25 kilowatts up to 500 kilowatts, and for wavelengths from 8,000 to 20,000 meters. The rotational speed of these machines, varying from 2500 revolutions per minute for the 500-kilowatt alternator to 6000 revolutions per minute for the 25-kilowatt alternator, was maintained rigorously constant by a very fast-acting regulator, whatever the load variations to which they might be subjected. Rotating in a rarefied atmosphere to avoid ventilation losses, energetically cooled by a very ingenious pressurized oil circulation, the high-frequency alternator offered a very high efficiency: more than 84 percent for the 500-kilowatt machine. As for the operation of these alternators, it was extremely simple. As soon as the final realization of the high-frequency alternator was completed, the Société Française Radioélectrique concerned itself with adding to the remarkable qualities of the machine (i)The engineers who especially contributed to this great success are Messrs. J. Bethenod and Marius Latour, at the Société Française Radioélectrique, Messrs. Roth, Belfils and Bilieux, at the Société Alsacienne de Constructions Mécaniques. The special regulator was the work of Mr. Thury.