The relations between the distortion caused by the action of analog level compressor and its recovery time, compression ratio, type of control circuit and the nature of control element are discussed. It becomes clear that the amount of the distortion strongly depends upon the recovery time of control. The results of calculations, when the attack time of control is supposed zero, give good approximations to the 3rd harmonic distortion measured. When compression is low, the distortions over the 5th harmonic, which are calculated under the supposed condition that the attack time is zero, have some difference from measured values. On the other hand, in the case of high compression, the calculated results become to give good approximation even at higher harmonics. When the actual recovery time is the same, the 3rd harmonic distortion is slightly affected by the type of control circuit, the nature of control element and compression ratio, but the distortion over the 5th harmonic is more affected by these conditions.
Computations are exerted using the finite element method for (1) SPL of the cones which vibrate as a piston with and without a dust-cap, and (2) the changes of SPL at the normal modes of vibration of the dust-cap, with and without the air loading, and are compared with the measured results. In computation, the air loading of the back space of the cone is neglected, and the air chamber inside the dust-cap is assumed to be closed. Results are as follows : (1) The large dip of the sound pressure response caused by the concave shape radiation of the cone is diminished due to the dust-cap. The agreement between the computation and the measurement is good. (2) A dust-cap attached to the part of the largest displacement of the normal mode of the cone affects that normal mode severely. Computed sound pressure peak level of that normal mode, not considering the air loading, is extremely smaller than the measured result. Computed result, the air loading considered, agrees well with the measured result. (3) The lowest normal mode of the dust-cap, in which the dust-cap vibrates almost in the same phase, causes a fairly large sound pressure peak level.
In this paper, the influence of the attraction of the magnet of the electro-magnetic transducer on a string vibration are studied with theoretical analysis making use of the measured magnetic attraction which is approximated by the static force and the negative stiffness component. The agreement between the computed and the photographed results of a string vibration is good. The results show that a string vibration has the inharmonicity which is a function of the affected position and the negative stiffness. Further, the inharmonicity results in beats whose frequencies are equal to f_n-nf_1, where f_n is the nth partial frequency. The beats are developed by the interference among the fundamental and the partials whose amplitudes are not so small as compared with fundamental. They are accentuated when n=2 or n=3.
The acoustic field around an elliptic cylinder located in the diffuse sound field is investigated. Especially the expected mean square values of sound pressure near a rigid strip, which is a special type of elliptic cylinders, are calculated in detail. The results show that the increase of SPL on the surface of a strip approaches to 3dB for sufficiently high frequencies. SPL along the central normal of the strip converges with oscillation to that original incident field as the observation point moves away from the strip. In the field more than one breadth of the strip apart from it, even in the direction of the normal, the scattering effect by the strip seems to be negligibly small. In order to verify theses theoretical results SPLs around a strip are measured in the reverberation chamber. The experimental values are in good agreement with the predicted ones. Moreover the results for a strip as above are compared with those for a disc the diameter of which is equal to the breadth of the strip.