Analyzer allows testing of polar response patterns of directional hearing aids

Abstract

INTRODUCTION It is easy to find glowing reports on the advantages of directional hearing aids. Just go online and ask Google for a list of directional hearing aid articles. There you will find scores of references to papers celebrating the virtues of directional hearing aids. These instruments improve the signal-to-noise ratio in difficult listening conditions. What does that mean to a person wearing directional hearing aids? He can better hear what he wants because interfering sounds, such as environmental noise and the speech babble from other conversations, are reduced. Directional hearing aids are designed to amplify speech and sounds that the wearer wants to hear while reducing all other noise. The benefits of directional hearing aids can be easily explained to the client, and satisfied customers are a joy to all. CLOUDS ON THE HORIZON This article provides a warning, however, that “directionality” is a delicate property that can be easily damaged, and it introduces a new tool that hearing health clinicians can use to determine if a hearing instrument is delivering the directionality for which it is designed. Two examples are given. The directional properties of the hearing aid rely on a delicate balance between the performance of a matched set of microphones, (usually two), and a system of signal delay(s) between them. If something occurs to disturb that balance, the directional properties of the hearing aid can be lost or upset. What types of problems can disturb the balance? People generate copious quantities of skin flakes and debris, some of which can collect in the microphone openings. Hair oil, sweat, excessive moisture, and hair spray all disturb or destroy the delicate balance between the microphones. Another problem in this new digital world is “forgetfulness”—not of the human wearer, but of the memory circuits that control the aid's operational parameters. These memory circuits that are programmed when the aid is fitted can forget what they were supposed to do. Figure 1 and 2 are polar plots of two digital hearing aids that use a popular industry chip that is built into several makes of hearing aids. Both of these hearing aids were originally programmed to exhibit directionality and moderate gain, and the client who wore them was an experienced hearing aid user capable of verifying if the directionality of the instruments was working. After they had been in use for 2 years, we ran a directional polar plot performance test at 2000 Hz at a level of 60 dB SPL with a new Fonix 8000 Hearing Aid Test System. As shown in Figure 1, the first hearing aid has completely lost its directivity. The second hearing aid, shown in Figure 2, still has a directional advantage, but in reverse, meaning that the sounds behind the client were being amplified more than the sounds in front of him. We examined these aids for physical problems and saw no sign that they had been abused. We found no microphone occlusion or dirt and observed no mechanical problems.Figure 1: Polar plot of an omnidirectional pattern from a “directional” aid.Figure 2: Polar plot of a reversed directional pattern.We then attempted to reprogram the instruments to correct the fitting. However, the electronic programmer box could not find them. It was as though they were not connected. The conclusion is that although the aids had been cared for, they simply forgot what their job was. One might call this problem “digital Alzheimer's disease.” If this testing had been done 6 months earlier, it is possible that the memory damage would have been correctable. Then the aids could have been resuscitated at the shop. But as they were, they had to be returned to the manufacturer for a rebuild. A good rule of thumb is to test directional aids for correct operation once or twice a year. More frequent testing may be needed if the patient and hearing aids are exposed to excessive warm weather, high levels of rain or humidity, excessive hair oil or hair spray, or substantial perspiration. HOW SHOULD YOU TEST DIRECTIONAL AIDS? Most hearing aid analyzers on the market today can be used to perform simple coupler measurements of gain or output to determine the provided amplification. Figure 3 shows a typical graph generated by a coupler test. This test was run at the level of 60 dB SPL with a modulated, broadband composite signal.Figure 3: Standard frequency response of a hearing aid.The directionality of a hearing aid is more difficult to test. Analyzer manufacturers have provided rudimentary tests of directionality for some time, but a clinical “polar plot” test of the directional pattern of the hearing aid was not available until recently. The new Fonix 8000 Hearing Aid Test System provides a special sound chamber with minimal acoustic reflections and a motor-driven mechanical rotator that can precisely turn the hearing aid in a 360° circle with respect to the sound source. Two types of tests are possible. The polar plot gives the gain or SPL from the hearing aid as a function of the aid's angle to the speaker. This plot may use either a single test frequency or a broadband signal. The difference test gives a frequency response of the difference between that measured with the aid pointed at the sound source and that measured at a chosen direction (usually at 180° or at the directional null). It is possible to generate a polar plot display for five frequencies simultaneously in about 35 seconds with one rotation of the aid for a five-frequency display. See Figures 4 and 5. These tests use a broadband signal. The figures shown are black and white, but the instrument's display is in colorFigure 4: Polar plot response of an aid with a noise-reduction feature and continuous drive signal.Figure 5: Polar plot of an aid with the signal interrupted between angular measures.The Fonix 8000 can also run all the coupler tests possible with the Fonix 7000. TESTING PROBLEMS If all hearing aids ran as linear amplifiers, testing a polar plot would be simple. However, rather complicated programming has entered the picture, as hearing aid manufacturers have attempted to allow the devices to compensate for changes in the user's environment and to achieve the best communication possible. These programs can affect measurement results. One approach to this problem was this one proposed by the ANSI hearing aid standards testing group: Allow the person testing the aid to change its internal program to remove all advanced features and have the aid act as a simple linear amplifier. Then parameters, such as gain, distortion, power output, and directional characteristics, can be tested to make sure the device's circuits, microphone(s), and receiver are operating properly. But in the clinical environment, it is often difficult or undesirable to alter the programming on a hearing aid that is in use. However, it is important to be able to run tests to be sure the device is operating correctly. Fortunately, many successful tests can be run on programmed hearing aids. Gain tests A measurement of gain with the hearing aid set to normal user settings is usually possible if a modulated, moderate level signal (such as the interrupted digital test signal available on the 8000) is used. On most hearing aids, an input level of 50 to 60 dB SPL will likely not substantially activate the AGC or expansion circuits of the hearing aid, and a modulated signal should prevent the noise-reduction algorithm from activating, thus allowing a valid test to be run. Figure 4 shows a polar plot test run at a level of 60 dB RMS with a continuously presented test signal. Plots at frequencies of 1000, 1500, 2000, 3000, and 4000 Hz are displayed. A difference test, the other form of directional plot, is displayed at the printout bottom on the right. The effects of noise reduction operation can be seen because the gains at 0° are slightly larger than those at 360°. The aid was equipped with noise-reduction capability and was in the process of turning back the gain at the end of the test. Now see Figure 5. Adjusting the signal so that it was interrupted between measurements stabilized the gain, resulting in a better match between the 0° and 360° measurements. Of course, the observant and knowledgeable operator recognizes the effect shown in Figure 4 and can use this knowledge to determine that the noise reduction is working, which can be a good thing. AGC effects Figure 6 shows the effects of AGC action. This test was run at the higher level of 80 dB SPL, and differences are seen between the polar plots of this figure and that of Figure 5. The AGC action removes some of the depth of the reverse gain loss seen in Figure 5. It acts to level out the aid's response. But the directional pattern is still apparent.Figure 6: Polar plot of an aid with AGC action at a higher test signal level of 80 dB SPL.Conclusion Problems have existed ever since directionality was introduced into hearing aids. These problems are caused by both the aided human's physical environment and the fragile digital circuits that control the hearing aid's directional properties. It has been impossible until now to fully test directional properties in the clinic because of the need for adequate acoustic environments and test equipment. However, the Fonix 8000 can perform these tests. It provides a mechanical test system and an adequate acoustic environment that can be used in the average clinical setting. In the future, we hope to perform further tests on a larger number of used directional aids. The goal is to get a better idea of the scope of the problem of directional response pattern alterations. Acknowledgment Many thanks go out to Dr. Robert Martin, whose advice is greatly valued. GLOSSARY OF TERMS Automatic gain control (AGC) AGC tries to keep the sound level from the aid constant, or more nearly so. When a directional aid is pointed away from the sound source, the sound intensity delivered to the aid's amplifier is reduced and the AGC action may increase the gain. This has the effect of filling in the reverse gain loss. Fortunately, most aids have the lower limit, or-kneepoint, of the AGC action under program control. This limit may be set to a moderate sound level of 50 to 60-dB SPL. Expansion Another program built into some instruments is called expansion. It is used to reduce background noise when the environment gets quiet. If the sound drops below conversational levels, the gain drops even farther. This program has the effect of increasing the difference between the forward direction and the reverse direction gains. Typically the knee level may be set around 40 to 50 dB SPL. Gain Gain is the calculated property of amplitude increase, or the output level divided by input. In dB, a logarithmic measure, gain is the output in dB minus the input in dB at each frequency. Noise-reduction algorithm As compared with a speech signal, which fluctuates, a continuous signal or noise band may be defined as noise. Modern digital hearing aids are sometimes equipped with a programming method (an algorithm) that causes the gain in one of the frequency channels to be reduced if the aid receives a continuous sound level in this channel. Polar plot Instead of using rectangular coordinates, some graphs are made in a circular format and are called polar plots. These graphs are a convenient way of visualizing the gain differences seen for directional microphones as a function of the sound presentation angle. Sound pressure level (SPL) SPL is the measure of the sound level of a signal, and is not a function of frequency. It is expressed in dB SPL, and is related to an absolute level that is quite low (.0002 microbar). This level once was considered to be the threshold of human hearing at 1000 Hz. Human hearing is often expressed in terms of dB HL, or hearing level. dB-HL is different for each test frequency.