Radio Broadcast (May 1929-Apr 1930)

Vibration Tests on Vacuum Tubes THE EXPERIMENTER'S ARMCHAIR By ROBERT S. KRUSE A motor-driven tube shaker used to produce vibrations of low frequency in testing microphonic vacuum tubes. 1AST month we seemed to be giving the transmitter the most consideration. A By way of evening up matters, we begin this time with a pure receiving story, having to do with the things that cause trouble in radio sets on airplanes and likewise in the ordinary home broadcast receiver. Mr. H. A. Snow, of the Radio Frequency Laboratories, Inc., Boonton, New Jersey, now has the floor. Vibration tests were made on several vacuum tubes to study microphonic noises produced in amplifiers when one or more of the tubes are jarred, or vibrated, by sound waves set up by loud speakers. These tests were made to determine at what frequencies in the audible range such microphonic noises occur and to determine the source of the noise. At that time (January, 1928) studies were made on only the 201a and the 222-type tubes. The method of study was as follows: The tube was mounted on a specially constructed vibrator unit adapted from a Western Electric loud speaker unit. The mounting of the tube was such that it received vertical and horizontal components of vibratory force produced when the vibrator unit was energized from a variable audio-frequency beatfrequency oscillator and power amplifier. The tube was connected as a detector in a typical two-stage amplifier such as used in broadcast receivers so that sounds produced in the amplifier by mechanical vibrations of the tube could be observed aurally by means of phones in the amplifier output. The tube under test was also observed visually by means of a magnifying lens system to determine just what element was vibrating and causing the sound in the amplifier. In examining a tube, the audio frequency supplied to the vibrator unit was slowly varied and the tube watched closely to observe element vibrations, and the sound produced by such vibrations observed by means of the head phones. Quantitative measurements of sound produced were not attempted, but the A vibrator unit employed in tests to determine the conditions under which vacuum tubes are microphonic. vibratory periods of various elements were measured accurately. It was foimd that elements of different tubes of the same type had different natural periods, as might be expected. However, the following series of natural periods of the 201a type tube was found to be representative: TUBE ELEMENT VIBRATING FREQUENCY direction (cycles) Plate and supports (as unit) Lateral 150 Filament and supports (as unit) " 280 One leg of filament *360 Second leg of filament *370 Plate — 1700 Grid (as unit) Lateral 330 *At normal filament voltage. These frequencies vary with filament temperature. At the frequencies shown above, each corresponding element would vibrate with considerable amplitude and produce a corresponding sound in the a.f . amplifier. The greatest amplitudes of vibration occurred laterally (considering tube held vertically). In addition to the vibrations listed above the individual grid wires each had natural periods which were difficult to observe separately but resulted in a considerable series of resonance vibrations in the neighborhood of 1000-2000 cycles. Also, there were many smaller vibrations of the elements in different modes that coidd be heard in the amplifier output but could not be observed in the tube with the crude lens system used. It was interesting to note that the natural period of the filament legs varied over a considerable range as the temperature was varied, due, no doubt, to the changing elasticity of the filament. TESTS WITH 222-TYPE TUBES Type 222 tubes were also examined, but in this case it was more difficult to attach a resonance frequency to one particular element because the tube structure is such that several elements are attached together at the top more or less rigidly so that at the natural period of any one element the entire system vibrated. There were, as a result of this, a number of combination natural periods, due not only to one element, but to the combination of several. The most pronounced amplitudes of vibration in the 222-type tube were at 130, 150, 220, and 500 cycles. The greatest amplitudes in most of the tubes examined were in a lateral direction; that is, with the tube held vertically, vibration of the elements occurred laterally. The conclusions reached from tests made are that the greater part of microphonic noises set up in tubes occurs from lateral vibrations of the various elements at their natural periods and that in the types of tubes studied, these vibration frequencies range from as low as 150 cycles to about 2000 cycles or higher, there being a large number of natural periods in each tube. Since so many natural periods exist in the audible range, it is a difficult matter when using a sensitive amplifier to prevent sound from a loud speaker being fed back into a tube and vibrate one or more of the elements at its natural period and so set up a howl, without radically changing the structure of a tube, or thoroughly shielding it from external vibrations by using some type of "shock-proof" mounting. The latter method is the only one available for sensitive amplifiers with the present tubes. In some recent broadcast receivers the effects of microphonic noise have been reduced greatly by eliminating one audio stage and using a high-voltage detector, thus reducing the audio amplification in the set which reduces the sound produced by tube vibration. It is interesting to hear the actual sounds produced when a tube is vibrated by means of the above unit over the audio range. The effect heard as the frequenoy is varied is as if the audio oscillator were producing its sound in a large hall made of sheet tin and half filled with tin cans, sheet metal, and other similar materials having a large number of resonant frequencies. For low-frequency vibrations there was also made a motor-driven tube shaker, which is shown in one of the accompanying pictures. An adjustable eccentric on the motor shaft and moveable sockets permit changes of stroke and the motor speed-control rheostat provides frequency change. The speed could be determined by comparing the pitches of • may, 1929 page 31 •