On extending audio feedback loops to include electrodynamic speakers
Beck, Earl R. (Earl Ray), 1916-, Jr
Pfeiffer, Paul E.
Master of Science
Summary. The purpose of the investigation was to determine the applicability of extending inverse feedback techniques to electrodynamic loudspeakers by finding first, what the general limitations were, and second, how much improvement in speaker characteristics could be secured. A simple lumped-constant representation of low-frequency speaker behavior was used for analysis of various types of electrical damping. Constants were determined for a typical speaker and mounting, and also for a suitable output transformer. The resulting combined equivalent circuit could be used to calculate the characteristics of the low-frequency cutoff, but was of little help for the upper cutoff. No simple, adequate representation of high-frequency speaker behavior could be found. An attempt was made to measure the frequency characteristics of the voice coil motional voltage by bridge measurements, but this was abandoned because of difficulties inherent in the method and lack of proper sized bridge elements. Three methods applicable to practical feedback systems were next used to obtain a speaker motional voltage. It was impossible to get usable high frequency data with two of the types, at least with the versions used, and the third gave misleading results. It appeared that the more obvious ways of obtaining a motional voltage were not as simple to use as one might expect from references in the literature. More accurate data obtained with a frequency modulation type of displacement pickup (whose output is much less convenient to use in a practical feedback system than the output of a velocity pick-up) put the upper limit to the use of feedback from 4 to 9 kc in a few typical cases, on a velocity to voice coil voltage basis, and with the latter held constant. Use of a high impedance amplifier would probably extend this limit at least one octave, but limitations in the measuring systems prevented finding the actual limit in this case. Transient response tests were made, and the results indicated qualitatively that electrical damping has noticeably less effect on speaker performance at the middle and high audio frequencies-than at the lower ones. This means that the beneficial effects of feedback would be less at the higher frequencies than at the low. Velocity-type feedback is potentially a useful means of controlling speaker damping and providing some distortion reduction in the speaker itself, at least in the range of piston action. Velocity-type feedback is a considerably more effective means of speaker damping than. the usual method of impedance reduction by feedback from the output of the amplifier. Further work is necessary to find the exact upper limit to the use of feedback with typical speakers, and to develop a velocity pickup which can cover all the usable feedback range. Frequency response and distortion tests on an actual speaker feedback system are needed to evaluate definitely the worth of such systems.