Radio Broadcast (Nov 1926-Apr 1927)

Record Details:

Something wrong or inaccurate about this page? Let us Know!

Thanks for helping us continually improve the quality of the Lantern search engine for all of our users! We have millions of scanned pages, so user reports are incredibly helpful for us to identify places where we can improve and update the metadata.

Please describe the issue below, and click "Submit" to send your comments to our team! If you'd prefer, you can also send us an email to mhdl@commarts.wisc.edu with your comments.




We use Optical Character Recognition (OCR) during our scanning and processing workflow to make the content of each page searchable. You can view the automatically generated text below as well as copy and paste individual pieces of text to quote in your own work.

Text recognition is never 100% accurate. Many parts of the scanned page may not be reflected in the OCR text output, including: images, page layout, certain fonts or handwriting.

588 RADIO BROADCAST APRIL, 1927 per second, and usually has its maximum below 1000 cycles. However, it has been found that a frequency range of about 50 to 5000 cycles is tolerable for natural reproduction of both speech and music.2 (see bibliography on page 590). Having considered the nature of speech and music, it is evident that the perfect reproducer should give constant response when actuated by constant audio signal impulses, and be free from resonant effects or hangovers of any sort. It is at once apparent that the action of the loud speaker may depend to some extent on the circuit elements used to couple it to the last amplifier tube, since the coupling devices may resonate the loud speaker at some audio frequency. It has been variously suggested that the quality of reproduced speech and music depends both upon the average response of the sound reproducing element to steady tones, and to the irregularity of the response frequency characteristics. The former serves as a basis for a rough estimate of quality and relative loudness of fundamental notes, while the latter is an index from which it is possible to predetermine the clarity of the reproduced speech and music. That is to say, a response-frequency characteristic obtained by applying steady single frequencies to the loud speaker gives only a general indication of the action of the loud speaker when it is actuated by transient notes, sudden stops, etc. To predict the effect of hangovers which tend to confuse the listener by changing the relative phase displacement of independent notes, it is necessary to know something about the resonant peaks in the response characteristic. Obviously, quality also depends on the nature of the load characteristic of the motor element. Since, as previously pointed out, the energy of both speech and music is more or less concentrated below 1000 cycles per second, the loud speaker, in actual service, must necessarily operate with a wide variation in amplitude. For this reason it is desirable that the efficiency of the loud speaker be approximately constant for all armature excursions commonly met in practice, otherwise, the large amplitude frequency notes, or the smaller amplitude high-frequency notes, will be over-emphasized. It has been shown elsewhere that musical tones between 200 and 2000 cycles have in general the same average intensity and that the human ear at the intensities generally used in reproduced music or speech has about the same sensitivity over this range.3 (see bibliography on page 590). Further, it has been shown that departures from faithful reproduction above and below this range are far less noticeable than departures within this range. For example, changes in response at 70 or 6000 cycles are about one tenth as serious as the same changes between 200 and 2000 cycles. As the range of maximum sensitivity is approached, given departures from true reproduction become more serious; thus, at 90 or 4000 cycles, a given departure from the true signal is about half as serious as at 1000 cycles. Consequently, if it is taken for granted that equal departures from the original signal do equal damage at points of equal auditory sensitivity, it is possible to arrive at a basis for determining the maximum departure of response from perfect reproduction allowable for tolerable reproduction. Using this as a basis, a response-frequency characteristic can be drawn which represents approximately the limit beyond which it is unnecessary to go for acceptable reproduction. Such a curve is shown in Fig. 4. From the standpoint of relative loudness, the response-frequency characteristic of any loud speaker which falls within the shaded area is substantially as acceptable as a perfect ■ ; A n-tplitbU Reproduce FREQUENCY FIG. 4 unit. This, of course, holds only for relatively smooth characteristics, since resonance peaks of relatively large amplitude and sufficient sharpness always introduce sustained vibrations or hangovers, and these constitute an entirely different type of distortion. Representative response characteristics of two commercial cone loud speakers, are shown in Fig. 5. A comprehensive method of measuring or 400 800 2000 4000 8000 CYCLES PER SECOND FIG. 5 rating a loud speaker in terms of its resonance peaks is probably as unnecessary as it is difficult. There are, however, a few conclusions that may readily be drawn from a first inspection of the response curve. A broad peak indicates high damping action, and a sharp peak, low damping. Therefore, if the broadness of any resonance peak is rated in terms of multiples or sub-multiples of the geometric mean frequency, the relative FIG. 6 length of the hangover can be approximately determined. Thus, a resonance peak of unit breadth at, say, seven tenths of its maximum height, will allow vibrations to persist roughly half as long as will one of the same amplitude and half its breadth, and in the former, these vibrations, for a given frequency, are about half as serious. It therefore appears worth while to examine closely the height and breadth of each resonance peak if its amplitude exceeds the average characteristic by 40 to 50 per cent. The absolute height of peaks that may be neglected is more or less a matter of opinion, and therefore, limits can only be fixed by measurement of the undesirable effects produced. Certainly, greatest harm will be produced by peaks falling within the frequency spectrum, most important from the standpoint of auditory sensitivity. Whence it would appear that the harmful effects of peaks can be weighted in accordance with the response curve of Fig. 4. THE LOUD SPEAKER'S MECHANISM HAVING considered, in general, the desirable characteristics of loud speakers, let us investigate the mechanism that is to produce these characteristics. A loud speaker may be considered as made up of three systems: 1. The motor element, which converts the electrical impulses into corresponding mechanical vibrations. 2. The coupling system, which transmits the mechanical vibrations from motor to diaphragm. 3. The diaphragm or loading device, which radiates the mechanical vibrations into the air as waves of sound. The simplest type of motor element in common use is the vibrating reed type. It consists, in general, of an armature pivoted or hinged at one end and actuated at the other, a magnet to supply a steady uni-directional magnetic flux, and a winding coupled to the magnetic circuit which is capable of superimposing an alternating or pulsating signal flux on the steady flux already present. A schematic diagram of such a unit is shown in Fig. 6. The free end of the armature is normally at rest in a steady unidirectional magnetic field supplied by the permanent magnet. When an alternating or pulsating current is passed through the coil coupled to the armature, the free end of the armature is alternately attracted and repulsed by the remaining pole. Thus, the electrical impulses in the coil are converted into mechanical vibrations which are, in turn, transmitted to a diaphragm through the medium of the driving rod. A brief inspection of the figure will show that this type of motor is not free from distortion. Let $ represent the steady flux across the air gap; a sin a>t, the flux due to a signal current through the windings, and P the force acting on the armature. Then: P = K (<j> + a sin tot)2 = 2K 4> a sin <ot + K a2 sin2 tot + K <]j2 , • K a2 , (2<$r + a2) = 2 K. 6 <* sin <ot cos 2 tot + K. 2 2 Obviously, the first term represents a reproduction of the signal impulse, while the second term represents a second harmonic of the signal4 (see bibliography on page 590). The remainder of the force adds a steady component to the steady pull exerted by the permanent magnet. It would seem, since that part of the coefficient of the second harmonic which is proportional to the signal, appears as a squared term, that the second harmonic could be made neg