助聽器晶片及系統---子計畫七：微機電式聲學元件暨助聽器異質整合(I)

Abstract

麥克風在助聽器中扮演著接收訊息的角色，如果有較強的訊號雜訊比 (SNR) 以及音源 定位能力，將會提升助聽器的品質。在本計畫中，首先利用仿生式的音源定位機制過濾 出訊息的來源方向，經由電路的運算，將四周的雜音降低，進而提聲音源訊號之SNR。 而所開發之仿生式麥克風除結構上將以環形支撐薄膜機械耦合式設計為出發點，在提升 元件自身之頻率響應範圍、靈敏度、方向性等指標同時，亦將嘗試於結構體上導入聲波 導管設計，期使達到機械式分頻進而省電之目的。此外本計畫亦將開發電磁式的微型喇 叭，此元件可以避免壓電式及靜電式喇叭對於高驅動電壓之運作需求。由於其在電聲功 率轉換上因較壓電式及靜電式設計為佳，更將使得喇叭於微瓦級功率消耗上需求較易達 成。由於電磁式的微型喇叭於製作上，必需利用非CMOS 相容之磁性材料製程以提升電 聲功率轉換之效能，而此舉將降低未來實現系統單晶片之可能性，因此本計畫在微型喇 叭製作開發上，亦將利用本實驗室所開發之奈米複合材料製程技術，藉由調變傳統 CMOS 中金屬導線之物理特性如同時具有高導電度與高磁性密度，使得其電磁特性得以滿足計 畫之所需，而此開發之製程終將因其與 CMOS 製程相容之特性，使得未來微型喇叭之製 作可由現今晶片製造技術來實現，進而達成系統單晶片之最終目標。本計畫亦將加入特 殊的設計如陣列聲道來增加頻寬並同時降低低音諧振點，並於製造微型喇叭中導入本實 驗室所開發之前瞻晶片接和技術，於喇叭結構上加入聲波導管的設計，進一步減少功率 的損耗與同時提升喇叭音質之可行性。最後本計畫希冀將利用系統封裝之架構以及計畫 中所預計開發之聲子晶體子並搭配計畫二所提出之耳道聲學模型以完成類比介面電路 系統與麥克風及喇叭之最後整合並達成微小化、低功耗、低回音與低雜訊之最終目的， 進而真正開發出低成本、高效能之 ITC 或 CIC 等款之耳道型助聽器。

This proposal presents our approaches to develop high performance acoustic devices including MEMS typed biomimetic microphone, electromagnetic driven microspeaker with acoustic wavehguide, and phononic bandgap crystal for the development of low powered ITC/CIC typed hearing aid. Based on silicon micromachined techniques, a silicon substrate will be micromachined as a needle shaped carrier on which four to five microspeakers will be fabricated on the top silicon carrier and connected with the waveguided structures These microspeakers are designed with different membrane sizes in diameter for different output pitches. On the other end of silicon carrier like a platform, there will be two spaces to host signal processing chips and battery using flip-chip bonding technique. Finally, the edge of platform will be connected to another silicon substrate like a folded silicon carrier using flexible Parylene cables in which all electrical interconnects to microphone arrays are directly fabricated on the carrier. The microphone based on the Ormia ochracea and Oxalis design. could have the capabilities of omni-directional sound response, less than 10° spatial resolution for noise reduction, around 10mV/Pa sensitivity, and 100 dB dynamic range, the microspeaker with waveguided structure would have the characteristics of ~1Vdriving voltage, 100dB SPL at 1KHz output power, and <500μw at 1KHz power consumption only with the size of 216mm3, and the phononic crystal with the mixture of diamond rod and fluid could have -30dB transmission in the range of 20Hz and 20KHz. Meanwhile via the developed integration technologies, like Au-Au thermal compressive Flip Chip bonding, we can integrate analog sensing and driving circuits and the developed audio components to form a hearing aid system onto a 「single」 chip and make such a system with the performance as well as or even better than the SOC. The primary objectives of the proposed research are: The primary objectives of the proposed research are: 1) the development of a high performance biomimetic microphone with sound source localization and noise reduction abilities, 2) the development of a low power, low distortion microspeaker, 3) the design and fabrication of phononic bandgap crystal to eradicate echo problem, 4) the development of heterogeneous integration technology using silicon carrier to package the whole hearing aid system, and 5) testing and physical modeling of the acoustic components for optimum analog circuit design. Since the final hearing aid system is aimed for the users speaking Chinese, it is our belief that the developed technologies will be very useful to the medical instrument manufacturers in Taiwan at leading stage.