The Performance and Theory of Loud Speaker Horns

(1) Experimental: By means of a vacuum tube oscillator to operate a loud speaker unit and a calibrated condenser transmitter, resistance-coupled amplifier, and a thermocouple and galvanometer, tests have been made at various frequencies to obtain the sound output from straight conical horns. The horns were made of heavy galvanized sheet iron and include lengths from 30 cm. to 305 cm. and final openings from 5 cm. to 90 cm. The results are plotted to show the sound pressure in dynes at the various frequencies up to 4,000 cycles. Marked resonance was obtained at all the lower frequencies and the effect was still observable up to 4,000 cycles. The resonant peaks due to the horns at the lower frequencies are frequently as high as three or four times the response at the adjacent anti-resonant frequencies. The ratio approaches unity for all horns at the higher frequency range and for all the horns with the large openings there are not large resonant peaks. At the peaks the resonant is most marked for the horns with small openings, and the "valleys" are not "filled in." The larger solid angle, however, causes sufficient resonance at all frequencies above the fundamental to fill in the valleys and to cut off the sharpness of the peaks. The result is a more uniform sound output. For the horns of each length and varying solid angle there is a particular solid angle and opening which give the optimum sound output.