Acoustic feedback in hearing aids occurs when the receiver of the hearing aid produces an acoustic signal that leaks to the microphone and causes oscillations and instability, limiting the hearing aid gain. The problem is worse in vented hearing aids, which are commonly used to reduce the unnatural feeling caused by the ear occlusion. Common solutions to the feedback problem are low-pass filtering for analog hearing aids, and adaptive feedback cancellation for digital hearing aids. Hearing aids are commonly designed primarily to compensate for the hearing loss, and therefore often suffer from feedback instability. In this work it is shown that the hearing aid can perform better if both the feedback problem and the required compensation are considered in the hearing aid design. An experiment was performed to study the hearing aid feedback mechanism in both an ear simulator and in human ears under various conditions. Then hearing aid filters that minimize the feedback problem for the human-ear data, while providing maximum gain for hearing loss compensation, were computed. The filter design problem was formulated as a convex optimization problem and solved using sequential quadratic programming. Results of the feedback path measurements and the hearing aid filters will be presented.
Read full abstract