Overview
Anechoic chambers are essentially sound proof rooms designed to simulate the effects of free space. In the theoretical free space, sound waves propagate from the source towards infinity with no returning waves. Unfortunately, this does not happen even within the best designed anechoic chambers. Very small reflections will always occur.
The International Standards Organization (ISO) has established numerous tests and guidelines which quantify the level of reflections that are acceptable for an anechoic chamber. One of the stringent requirements to perform any of these qualifying tests is the presence of an omnidirectional sound source. As its name implies, an omnidirectional sound source is a source (loudspeaker, Helmholtz resonator, etc.) that emits single frequency spherical waves evenly in every direction. In other words, the sound pressure level is only a function of the radial distance from the source.
The purpose of this research is to develop an omnidirectional sound source which covers the entire audible frequency range (20Hz-20kHz).
Test Setup &
Procedure
For a sound source to be deemed omnidirectional, it must meet the requirements set forth in ISO 3745 (Annex A). This document specifies to first place the test source in the geometric center of the chamber. As the source emits a pure tone at the desired frequency, a total of 56 sound pressure levels must be recorded at a distance of 1.5m from the source. More specifically, the phi(f) or azimuth angles (parallel to walls) of the microphone locations must be 20, 40, 60, 80, 100, 120 and 140 degrees. The theta(q) or polar angles (parallel to floor) of the mic locations must be 0, 45, 90, 135, 180, 225, 270 and 315 degrees. This results in a subsequent total of 56 measurements (7 azimuth X 8 polar angles).
In order to avoid having to use 56 microphones, a semicircle with 1.5m radius was created out of PVC piping. As shown in Figure 1, a total of seven random incidence microphones are evenly spaced along the arc.
The source is then suspended in the center of the semicircle and rotated about itself by 45 degree increments per ISO 3745 requirements. This procedure is repeated for every frequency of interest.
Figure 1. Microphone arc within anechoic chamber
Data Processing & Results
During each source rotation increment, the 1/3-octave sound pressure level containing the frequency of interest is recorded. Ideally, all seven microphones would record the identical sound pressure level at every source angle. This would result in a perfeclty onmidirectional source. However, this is usually not the case. But as long as any deviations fall within the specifications of ISO 3745, the source can still be considered omnidirectional for purposes of ancechoic chamber qualification. The frequency based deviation limits are listed in Table 1.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
An example
of a test source is shown in Figure 2. It is composed of six (three
shown) 4" woofers each attached to the face of a cube made
of medium density carbon fiberboard. All speakers are attached
in parallel to one function generator. This particular was suspended
from the ceiling of the chamber using a series of hooks. Rotation
of the source about the polar axis was therefore very easy. Plots
of the effectiveness of this source are shown in Figures 3 and
4. In an effort to avoid clutter, the resolution of the polar
graph has not been included. Each solid line equals a deviation
of 5dB with the middle line being the mean deviation. The dotted
lines represent the tolerance linmits. Therefore, anything within
these dotted lines represents an acceptable sound pressure measurement.
Note that at higher frequencies, this source design becomes ineffective
as its deviations exceed the tolerance limits of ISO 3745. However,
it proves to be a very omnidirectional source below 200Hz.
Figure 2: Omnidirectional Box Source
Future Work
Although the box design seems adequate for low frequencies, it does not cover the entire frequency range needed. For frequencies up to 1000Hz, two 4" woofers bolted face to face seems to be effective. Again, however, it begins to fail above 1kHz. Therefore, for the highest frequencies, a compression driver with a long narrow rod attached to it may prove to be effective. The theory is that all the sound will propagate through the tube and exit the rod opening, thereby acting like an omnidirectional monopole. The reason the rod needs to be narrow is that the characteristic length of the opening needs to be smaller than the wavelength of the emitted pure tone. At high frequences, this is a very small dimension. This would also serve to explain the reason why the relatively large box source is only effective at very low frequencies.