Radio Broadcast (May 1928-Apr 1929)

21 RADIO BROADCAST resonant effects but as soon as the sides are bent around, as in a cabinet, bad resonance occurs. Part of this resonance is due to the sides vibrating and part due to the natural period of the cavity. Making the cabinet of heavy wood helps reduce the resonating effect due to the sides vibrating. That this effect of cabinet resonance is very bad can be shown experimentally. The effect on music and speech is to change the quality greatly. It becomes booming in nature because the resonance is generally at low frequencies, and, in addition, the air chamber attenuates the high frequencies. Response Measurements MEASUREMENTS were made by means of a W. E. 387w transmitter and calibrated amplifiers over the audio-frequency band in an effort to learn something about cabinet resonance. The circuit arrangement used is shown in Fig. 4 in which a special beat-frequency audio oscillator was used as a source of sound. This audio oscillator had a range of from less than 30 cycles up to about 15,000 cycles and was, of course, continuously variable. Particular care was taken to minimize standing waves of sound in the room. The most practical method is to have the "mike" less than a foot from the loud speaker so that the direct sound wave is much stronger than the reflected waves. Numerous response curves were run with the "mike" in different locations. Some trouble was had from room resonance and re flecting surfaces since either the loud speaker or the "mike" had to be moved for the different runs. Even with these effects it is quite evident that cabinet resonance is pronounced as shown in curve b of Fig. 5. A larger cabinet generally has a lower resonant period, but because of audio amplifier deficiencies, it may not be very noticeable. Padding the inside of the cabinet with felt does not help much since felt is not an efficient absorbing material for low frequencies. Therefore, felt padding may attenuate the high frequencies more and tend to make the quality even more drummy in character. Felt padding helps occasionally in damping the sides to prevent vibration. Lining the cabinets with acoustic celotex or some such material should help greatly. Mounting the entire loud speaker unit in thick felt seems to remove the cabinet resonance but this cuts down the sound output nearly half. Only the front can emit sound in this case so a larger power tube is necessary to prevent overloading in the audio amplifier for the same sound output. Effect of Small Cabinets THE harmful effect of small cabinets on the higher frequencies is shown vividly in the curves c of Fig. 5. The solid curve was taken with the microphone about 15 centimeters in front of the loud speaker, the dotted curve was taken with the microphone at the same distance to the rear and the dot-dash curve was made with the microphone on one side. The dot-dash curve shows the effect of cabinet resonance since the "mike" was near one of the vibrating surfaces. The sudden drop at low frequencies is probably due to interference of sound waves emitted from the back and front of the loud speaker. The dotted curve d of Fig. 5 shows the effects of cabinet resonance and the attenuation of the high frequencies. Evidently the cabinet cavity acts like a condenser in absorbing more energy on the higher frequencies. It is like a horn loud speaker in which there is a large air cavity between the diaphragm and the throat of the horn. It is quite a wellknown fact that such a cavity attenuates the high frequencies greatly. If the air chamber or cavity is large enough with respect to the diaphragm, such as with a console cabinet, this attenuation of the high frequencies is of much less importance. If a small cabinet must be used, drilling a few large holes in the sides should help reduce both cabinet resonance and high-frequency attenuation. These holes would prevent the small cabinet from acting as a horn, but the effective baffle size would be diminished somewhat so the very low notes would be down a little in level. Large cabinets such as those used to completely house the radio receiver may reduce the resonance to a minimum by using a screen back for the cabinet and by not having any shelves inside of the cabinet. The use of a couple of strips of acoustic material, such as type BB Celotex, fastened to the sides or sides and top inside of the cabinet should make this form of cabinet practically as good as a flat baffle. Vertical Divisions = 5 DB Heavj Lines = 0 I B ' Withoi ( it Equalizei V *(A) With Equali /< k. i •■\ \ r Small C ibinet r \^ / Flat Baffle (B) \ 1 \ Front 6" \ V n H» t-X TT -P — V \ / 1 u * Side 6" . V Back 6" i t « • * — -V h • • \J i \ f (C) \ \ ► 1 o o o o o o o o o irj 00 FREQUENCY FREQUENCY Fig. 5 — Response curves of several dynamic-type loud speakers measured under different conditions. Curve A, Jensen loud speaker; curve B, Jensen loud speaker with different baffles {measured five feet in front of loud speaker); curves C, dynamic loud speaker in a small cabinet; curve D, small-cabinet-type dynamic loud speaker; curve E, Magnavox dynamic loud speaker; curve F, Jensen loud speaker with large cone and no filter or equalizer. • inarch, 1929 page 318 •