Radio broadcast .. (1922-30)

590 Iron losses are due to hysteresis and to eddy currents. Hysteretic effect may in general be sufficiently minimized by the use of a good grade of iron, and lamination of that part of the magnetic circuit traversed by the alternating flux appreciably reduces eddy current losses. Eddy currents are due to induction and are induced in the metallic portions of the motor structure by virtue of the changing magnetic flux. Eddy current losses increase as the square of the frequency, and hence tend to reduce the high-frequency response of the motor. Eddy current losses are also dependent on the excur- sions of the armature and thus may cause ampli- tude distortion due to their variation over the signal cycle. The frequency distortion due to eddy currents is, however, generally far more serious than the amplitude distortion, and their effect is generally quite apparent in response characteristics. In order that the armature vibrations be im- parted to the air as sound waves, it is necessary to couple a loading device to the motor element. The loading device may consist of a horn to- gether with a small diaphragm and air chamber, or of a large diaphragm which imparts the vi- brations directly to the air. The function of either type of loading element is much the same although their action differs somewhat. Properly designed horns apply an almost constant load to a motor element and may thus be made to produce a much smoother response-frequency characteristic than the cone type (large dia- phragm) device 6 (see bibliography on this page). The essential difference in the two lies mainly in the fact that the load presented by the horn is almost pure radiation resis- tance over the operating range, while that supplied by the cone is far from constant, re- sulting generally in an irregular response char- acteristic, and is very similar in its action toward the mechanical system to a complex impedance load in an electrical circuit.f Aside from the dif- ference in relative smoothness of their response characteristics the frequency band covered in the two cases is quite different. See Fig. 11. The lowest frequency radiated by a horn is a function of its length. Many horns function in a manner similar to an open organ pipe, in that the lowest frequency transmitted is F = VEL. OF SOUND IN AIR 2 (LENGTH OF HORN) to a first approximation. This obviously depends on the shape of the horn and the shape of the opening. The lowest frequency efficiently radi- ated by the cone is, among other factors, de- RADIO BROADCAST tThis applies to large flexible diaphragms such as paper cones. With a very stiff diaphragm of small size, plunger action and more uniform response may result. FIG. I I termined by the size of the structure and the amplitude of motion. In addition, the highest transmitted frequency efficiently radiated by this type of loading device is partly determined by the effective high-frequency mass of the dia- phragm. In some diaphragms, at high frequen- cies, only a small portion of the diaphragm near the driven point is effective as a radiator, the remainder of the device acting in general as a power absorbing network. That is to say, the high-frequency vibrations may travel from apex to edge of the cone as well as being directly radi- ated from the driven point. The effect of these impulses traveling across the face of the dia- phragm is twofold: First, power lost in deforma- tion of the diaphragm face, and, second, out-of- phase radiation. These may frequently occur to such an extent that the radiated sound is much reduced. In addition, the usual diaphragm has many resonant points of its own and thus may present a variable load to the motor. The efficiency and response-frequency char- acteristics of a loud speaker depend to a large extent upon the device used to couple the motor to the loading element. In the case of the horn radiator, this device usually consists of a short light driving rod and is, in general, quite ef- ficient and relatively free of distorting effects. In the case of the large diaphragm (cone) loading device, the coupling device must in general in- clude a mechanical transformer, which amounts to a lever or system of levers for increasing the force. In the ideal case the mechanical trans- former consists of a frictionless lever of zero mass. In practical cases it may consist of mem- bers having appreciable mass and considerable stiffness of suspension, amounting to a complex network, which further complicate the action of the loud speaker as a unit. This particular phase of the subject, however, has been treated in detail elsewhere 7 (see bibliography on this page). The human ear which in the end is the final judge of quality, is far from a perfect instrument. Its response characteristic is far from linear either with frequency or amplitude 8 (see biblio- graphy on this page). Moreover, the ear is in APRIL, 1927 itself a modulator, due to the nonlinearity of its characteristics 9 (see bibliography on this page). Such being the case, it is often possible for an observer to apparently hear a fundamental note, when only harmonics of the note actually im- pinge on the ear drum 10 (see bibliography on this page). Therefore, it may often be permis- sible to allow a certain amount of distortion to actually take place, and still maintain tolerable quality. Thus, harmonics produced by a loud speaker may often cause a slight apparent in- crease in efficiency without materially affecting quality. The naturalness of reproduced speech and music depends to a large extent on the energy level at which it is reproduced. That is to say, the psychological reaction of the ob- server depends on whether or not the signal is delivered at normal intensity. Again, harmonics of the original signal are frequently not detected by the ear until the energy level is such that the harmonic output of the reproducer approaches that of the original signal. Moreover, relatively large irregularities in the response characteristic of an acoustic reproducer are frequently allowable at either end of the frequency spectrum, as pre- viously • pointed out. Consequently, a perfect reproducer, having a linear response characteris- tic, might not appear to have an appreciable advantage over a less perfect device with a reasonably acceptable characteristic. 8—' 9— BIBLIOGRAPHY -"Nature of Language," R. L. Jones. Electrical Com- munication —Vol. 2, No. i. July, 1923. -"High Quality Reproduction of Speech and Music." W. H. Martin and H. Fletcher. Journal A. 1. E. E. Vol. XL!I, No. 3. March, 1924. -"The Relation between the Loudness of a Sound and Its Physical Stimulus," J. C. Steinberg. Physical Review. October, 1925. -"Design of Telephone Receivers for Loud Speaking Purposes," C. R. Hanna. Proc. I. R. £. August, 1925. "Theory of Magneto-Mechanical Systems," R. L. Wegel. Journal A. I. E. E. October, 1921. "The Electromagnetic Theory of the Telephone Re- ceiver," A. E. Kennelly and H. Nukiyama. Proc. A. I. E. E. April, 1919. "Fouriers Series and Spherical Harmonics," Byerly. -" Function and Design of Horns for Loud Speakers," C. R. Hanna and J. Slepian. Journal A. 1. E. E. February, 1924. "Acoustical Impedance and Theory of Horns and of the Phonograph," A. G. Webster. Proc. National Acad- emy of Sciences. 1919. Vol. 5. " Experiments on the Vibration of Air in Conical Horns," A. G. Webster. Proc. National Academy of Sciences. 1920. Vol. 6. "Theory of Sound," Lord Rayleigh. Vol. 2. " Performance and Theory of Loud Speaker Horns," Goldsmith and Minton. Proc. I. R. E. August, 1924. •"Methods of High-Quality Recording and Reproducing of Music and Speech Based on Telephone Research,' J. P. Maxfield and H. C. Harrison. Bell System Tech- nical Journal. July, 1926. 'Physical Measurements of Audition and Their Bear- ing on the Theory of Hearing," Harvey Fletcher. Journal of Franklin Institute. Vol. 196, No. 3. Sep-_ tember, 1923. 'Auditory Masking and Dynamics of the Inner Ear," R. L. Wege! and C. E. Lane. Physical Review. Febru- ary, 1924. 'Physical Criteria for Determining the Pitch of a Musical Tone," H Fletcher. Physical Review. September, 1923. RADIO BROADCAST is the official publication of the Radio Ciub of America, through whose courtesy, the foregoing paper has been printed here. RADIO BROADCAST does not, of course, assume responsibility for controversial statements made by authors of these papers. Other Radio Club papers will appear in subsequent numbers of this magazine