200929170 九、發明說明： 【發明所属之技術領域】 本發明係有關於一種音源分析裝置，特別係有關於一 種具有收音單元以及轉換單元之音源分析裝置。 【先前技術】 傳統的語音辨識技術通常是先蒐集足夠的語音樣本， 並經過擷取適當的語音特徵之後，進而建立聲學模型所需 ❹ 的參數。一般執行語音辨識的程序大致為：先透過收音設 備接收語音訊號，接著利用適當的演算法擷取出適當的語 音特徵，再與資料庫中的模型進行比對。由於目前處理語 音訊號的處理方式大多必須透過軟體或是繁複的硬體電路 來加以實現，因此如何增進聲音取樣分析之效率，並且透 過簡化機構設計以達到微型化之目的始成為一重要之課 題。 ❿【發明内容】 本發明一實施例中揭示一種音源分析裝置，用以分析 一音源，上述音源包括一第一聲波，上述音源分析裝置包 括一收音單元以及一轉換單元，其中收音單元包括一第一 共振結構體及一第一壓阻元件，上述第一壓阻元件設置於 第一共振結構體上。前述轉換單元包括一發射/接收區以及 一第一反射區，其中第一反射區連接第一壓阻元件，當發 射/接收區產生一發射聲波至第一反射區且第一共振結構 200929170 體可與第一聲波產生一共振現象時，第一反射區根據發射 聲波與第一壓阻元件產生一第一表面聲波至發射/接收 區，前述發射/接收區則根據第一表面聲波得到第一聲波之 頻率。 於一實施例中，前述收音單元更包括複數個第一共振 結構體以及複數個並聯之第一壓阻元件，前述第一壓阻元 件設置於第一共振結構體上。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound source analyzing device, and more particularly to a sound source analyzing device having a sound collecting unit and a converting unit. [Prior Art] Conventional speech recognition technology usually collects enough speech samples first, and then extracts appropriate speech features to establish the parameters required for the acoustic model. Generally, the procedure for performing speech recognition is roughly as follows: first, the voice signal is received through the radio device, and then the appropriate speech feature is extracted by using an appropriate algorithm, and then compared with the model in the database. Since most of the processing methods for processing voice signals must be implemented by software or complicated hardware circuits, how to improve the efficiency of sound sampling analysis and to achieve miniaturization by simplifying the design of the mechanism has become an important topic. The present invention discloses a sound source analyzing device for analyzing a sound source, the sound source includes a first sound wave, and the sound source analyzing device includes a sound collecting unit and a converting unit, wherein the sound collecting unit includes a first a resonant structure and a first piezoresistive element, wherein the first piezoresistive element is disposed on the first resonant structure. The foregoing conversion unit includes a transmitting/receiving area and a first reflective area, wherein the first reflective area is connected to the first piezoresistive element, and the transmitting/receiving area generates an emitted sound wave to the first reflective area and the first resonant structure 200929170 is When the first sound wave generates a resonance phenomenon, the first reflection region generates a first surface acoustic wave to the transmitting/receiving region according to the emitted sound wave and the first piezoresistive element, and the transmitting/receiving region obtains the first sound wave according to the first surface acoustic wave. The frequency. In one embodiment, the sound receiving unit further includes a plurality of first resonant structures and a plurality of first piezoresistive elements connected in parallel, and the first piezoresistive elements are disposed on the first resonant structure.
於一實施例中，前述發射/接收區具有一發射/接收電 極，且第一反射區具有一第一反射電極，其中發射/接收電 極以及第一反射電極為交指狀電極。 於一實施例中，前述發射/接收電極以及第一反射電極 具有鋁(A1)材質。 於一實施例中，前述第一共振結構體包括一基材以及 一第一懸空薄膜，其中基材與第一懸空薄膜形成一第一腔 艚。 於一實施例中，前述基材為碎基材。 於一實施例中，前述第一懸空薄膜大致呈矩形或圓形。 於一實施例中，前述第一懸空薄膜及第一腔體由微機 電製程所形成。 於一實施例中，前述轉換單元更包括一基板，其中發 射/接收區以及第一反射區形成於前述基板上。 於一實施例中，前述基板含有氮化鋁（A1N)、氧化鋅 (ZnO)、鈮酸鋰（LiNb03)、钽酸鋰（LiTa03)或者锆鈦酸鉛 (PZT)材質。 200929170 於一實施例中，前述音源更包括一第二聲波，轉換單 元更包括一第二反射區，且收音單元更包括一第二共振結 構體及一第二壓阻元件，其中第二壓阻元件設置於第二共 振結構體上。當發射聲波傳遞至第二反射區且第二共振結 構體可與第二聲波產生一共振現象時，第二反射區根據前 述發射聲波與第二壓阻元件產生一第二表面聲波至發射/ 接收區，發射/接收區根據第二表面聲波可得到第二聲波之 頻率。 @ 於一實施例中，前述收音單元更包括複數個第二共振 結構體以及複數個並聯之第二壓阻元件，其中第二壓阻元 件設置於第二共振結構體上。 於一實施例中，前述第二反射區具有一第二反射電 極，其中第二反射電極為交指狀電極。 於一實施例中，前述第二共振結構體包括一基材以及 一第二懸空薄膜，其中基材與第二懸空薄膜形成一第二腔 體。 Ο 為使本發明之上述目的、特徵、和優點能更明顯易僅， 下文特舉較佳實施例並配合所附圖式做詳細說明。 【實施方式】 首先請參閱第1A圖，本實施例之音源分析裝置主要包 括一收音單元Μ以及一轉換單元T，其中該轉換單元T例 如為一交指叉式表面聲波換能器（Inter-Digital Transducer, IDT)，包括一基板B、一發射/接收區e、一第一反射區rl 8 200929170 以及一第二反射區r2，前述收音單元Μ則包括一基材S、 一第一壓阻元件P1以及一第二壓阻元件P2。如第1A圖所 示，前述發射/接收區e以及第一、第二反射區rl、r2皆位 於基板B上，在發射/接收區e内設有一發射/接收電極E， 在第一、第二反射區rl、r2内則分別設有一第一反射電極 R1以及一第二反射電極R2,其中發射/接收電極E以及第 一、第二反射電極Rl、R2可採用交指狀電極。 接著請一併參閱第1A〜1C圖，第一壓阻元件P1設置 ❹ 於基材S上方之第一懸空薄膜F1，並且在第一懸空薄膜 F1與基材S之間形成一第一腔體C1(如第1B圖所示），其 中第一懸空薄膜F1與第一腔體C1可組成一第一共振結構 體；此外，前述第二壓阻元件P2設置於基材S上方的第二 懸空薄膜F2，並且在第二懸空薄膜F2與基材S之間形成 一第二腔體C2(如第1C圖所示），其中第二懸空薄膜F2與 第二腔體C2可組成一第二共振結構體。於本實施例中， 前述第一、第二壓阻元件P卜P2係分別與第一、第二反射 ❹ 電極R1、R2耦接並分別組成一第一迴路L1以及一第二迴 路L2。需特別說明的是，前述第一、第二共振結構體的共 振頻率係可由基材S、懸空薄膜FI、F2以及腔體Cl、C2 的尺寸大小、材質以及厚度所決定。 如第1A圖所示，由於前述第一、第二共振結構體所對 應的共振頻率不同，因此當接受到外界音源W中不同頻率 的聲波擾動時，可藉由第一、第二懸空薄膜F卜F2上之第 一、第二壓阻元件P卜P2分別感測出相異基頻的震動，並 9 200929170 可在第一、第二迴路Ll、L2中產生相對應之等效電阻負 載。於操作時，可輸入一交流電訊號給發射/接收區e中的 發射/接收電極E，使其在基板B上產生一發射聲波V(大於 20K Hz)至右側第一、第二反射區r卜r2内的第一、第二 反射電極R1、R2 ;當發射聲波V傳遞至第一反射區rl且 第一共振結構體可與音源W之一第一聲波產生共振現象 時，第一反射電極R1會根據前述發射聲波訊號V與第一 壓阻元件P1產生一第一表面聲波VI至發射/接收區e ;同 © 理；當發射聲波V傳遞至第二反射區r2且當第二共振結構 體與前述音源W中之一第二聲波產生共振現象時，第二反 射電極R2則會根據前述發射聲波訊號V與第二壓阻元件 P2產生一第二表面聲波V2至發射/接收區e，其中第一、 第二聲波之頻率互異。 再請參閱第1D圖，該圖表示發射/接收電極E所接收 到之第一、第二表面聲波VI、V2回波損耗(return loss)之 時域波形，其中由於第一、第二反射電極Rl、R2之電阻 ❹ 負載不同，導致第一、第二表面聲波VI、V2所分別對應 之回波損耗(return loss)蜂值Ql、Q2大小相異，藉此可得 知音源W中第一、第二聲波之頻率及其強度。 如前所述，透過發射/接收區e可傳送發射聲波V至第 一、第二反射區rl、r2，接著可根據第一、第二反射電極 Rl、R2所傳回之第一、第二表面聲波VI、V2而分別得知 音源W中第一、第二聲波的頻率與其對應的強度大小，其 中前述發射/接收電極E更可連接至一控制單元（未圖示）， 10 200929170 透過控制單元可對前述第一、第二表面聲波VI、V2進行 訊號處理或者相關的頻域分析。 接著請參閱第2A圖，於另一實施例中的基材S上方 亦可分別設置複數個第一、第二懸空薄膜FI、F2以及複數 個第一、第二壓阻元件PI、P2，其中第一壓阻元件P1分 別對應地設置於第一懸空薄膜F1上，第二壓阻元件P2則 分別對應地設置於第二懸空薄膜F2。特別地是，前述第一 壓阻元件P1係相互並聯同時耦接第一反射電極R1，此外 ❹ 第二壓阻元件P2則相互並聯同時耦接第二反射電極R2。 如第2B〜2C圖所示，前述第一、第二懸空薄膜FI、F2下 方分別形成有第一、第二腔體Cl、C2，由於第一、第二懸 空薄膜FI、F2以及第一、第二腔體Cl、C2的大小尺寸相 異，因此可透過第一、第二懸空薄膜FI、F2上的第一、第 二壓阻元件P1、P2分別感測出上相異基頻的震動。需特別 說明的是，前述第一、第二懸空薄膜FI、F2以及第一、第 二腔體Cl、C2可針對不同的共振頻率而組成不同之共振 ❿ 結構體，其中共振結構體的共振頻率可由基材S、懸空薄 膜FI、F2以及腔體Cl、C2的尺寸大小、材質與厚度等參 數所決定。本實施例透過設置複數個並聯之第一壓阻元件 P1以及第二壓阻元件P2,不僅可達到多點接收的效果以提 升訊號/雜訊比（S/N ratio)，同時可避免因部分結構損壞而 導致整個音源分析裝置失效，進而可提升整體的效能與穩 定性。 舉例而言，前述轉換單元T所使用的電極和收音單元 200929170 Μ中的懸空薄膜F1、F2結構可透過微機電製程一併形成 於基板S上’藉以達到縮小系統尺寸以及微型化之目的， 其中前述懸空薄膜亦可為長條形的懸f結構所取代。此 外，前述基材S可為㈣玻璃等介電材料，前述發射/接收 電極E以及第-、第二反射電極Rb R2料採用銘(A1) 材質，至於前述基板^縣域電材料所組成，例如可採 用氮化鋁（A1N)、氧化鋅(zn〇)、鈮酸鋰(LiNb〇3) 、组酸鐘__ (LiTa03)或者锆鈦酸鉛(PZT)等材質。 接著请-併参閱f 3A、3B 1]，於# 一實施例中的第 -、第二壓阻兀件ΡΓ、P2’係設置於基材s上之第一懸 空薄膜F1’ ’其中第—懸空薄膜F1’大致呈圓形且在其下 方形成有-第-腔體C1’。需特職明的是，由於前述第 -，、第二壓阻7G件PI’、P2’ >別設置於第―懸空薄膜 F1上的不同位置，因此可分別感測出第一懸空薄膜F1， 上的不關率震動’進而可透過前述轉換單元T得到音源 W中的不同聲波頻率極其強度。 綜上所述，本發明提供一種音源分析裝置，其中藉由 f二收音單元與—轉換單元（例如一交指叉式表面聲波換 月匕盗）搞合’可用以偵測外界音源中的不同的聲波頻率與強 度’不僅機構fl單同時可達到微型化之目的。 12 200929170 【圖式簡單說明】 第1A圖表示一音源分析裝置示意圖； 第1B圖表示第1A圖中沿Α1-ΑΓ之剖面圖； 第1C圖表示第1A圖中沿A2-A2’之剖面圖； 第1D圖表示第一、第二表面聲波之回波損耗時域波形 示意圖； 第2A圖表示本發明另一實施例之示意圖； 第2B圖表示第1A圖中沿B-B’之剖面圖； 〇 第2C圖表示第1A圖中沿C-C’之剖面圖； 第3A圖表示本發明另一實施例之示意圖；以及 第3B圖表示第3A圖中沿D-D’之剖面圖。 【主要元件符號說明】In one embodiment, the aforementioned transmitting/receiving area has a transmitting/receiving electrode, and the first reflective area has a first reflective electrode, wherein the transmitting/receiving electrode and the first reflective electrode are interdigitated electrodes. In one embodiment, the transmitting/receiving electrode and the first reflective electrode are made of aluminum (A1). In one embodiment, the first resonant structure includes a substrate and a first suspended film, wherein the substrate forms a first cavity with the first suspended film. In one embodiment, the substrate is a broken substrate. In one embodiment, the first suspended film is substantially rectangular or circular. In one embodiment, the first suspended film and the first cavity are formed by a microcomputer process. In one embodiment, the converting unit further includes a substrate, wherein the transmitting/receiving area and the first reflective area are formed on the substrate. In one embodiment, the substrate comprises aluminum nitride (A1N), zinc oxide (ZnO), lithium niobate (LiNb03), lithium niobate (LiTa03) or lead zirconate titanate (PZT). In an embodiment, the sound source further includes a second sound wave, the converting unit further includes a second reflective area, and the sounding unit further includes a second resonant structure body and a second piezoresistive element, wherein the second piezoresistive The component is disposed on the second resonant structure. When the emitted acoustic wave is transmitted to the second reflective region and the second resonant structural body can generate a resonance phenomenon with the second acoustic wave, the second reflective region generates a second surface acoustic wave to transmit/receive according to the foregoing emitted acoustic wave and the second piezoresistive element. The area, the transmitting/receiving area can obtain the frequency of the second sound wave according to the second surface acoustic wave. In an embodiment, the foregoing sound receiving unit further includes a plurality of second resonant structures and a plurality of parallel second piezoresistive elements, wherein the second piezoresistive elements are disposed on the second resonant structure. In one embodiment, the second reflective region has a second reflective electrode, wherein the second reflective electrode is an interdigitated electrode. In one embodiment, the second resonant structure includes a substrate and a second suspended film, wherein the substrate and the second suspended film form a second cavity. The above described objects, features, and advantages of the present invention will become more apparent from the detailed description of the preferred embodiments. [Embodiment] Referring first to FIG. 1A, the sound source analyzing device of the present embodiment mainly includes a sound receiving unit Μ and a converting unit T, wherein the converting unit T is, for example, an interdigitated surface acoustic wave transducer (Inter- The digital transducer (IDT) includes a substrate B, a transmitting/receiving area e, a first reflective area rl 8 200929170, and a second reflective area r2. The foregoing radio unit includes a substrate S and a first piezoresistive Element P1 and a second piezoresistive element P2. As shown in FIG. 1A, the foregoing transmitting/receiving area e and the first and second reflective areas rl and r2 are all located on the substrate B, and a transmitting/receiving electrode E is disposed in the transmitting/receiving area e, in the first and the first A first reflective electrode R1 and a second reflective electrode R2 are respectively disposed in the two reflective regions rl and r2, wherein the transmitting/receiving electrode E and the first and second reflective electrodes R1 and R2 can adopt interdigitated electrodes. Referring to FIGS. 1A to 1C, the first piezoresistive element P1 is disposed on the first suspended film F1 above the substrate S, and a first cavity is formed between the first suspended film F1 and the substrate S. C1 (as shown in FIG. 1B), wherein the first suspended film F1 and the first cavity C1 may constitute a first resonant structure; further, the second piezoresistive element P2 is disposed on the second dangling above the substrate S a film F2, and a second cavity C2 is formed between the second suspended film F2 and the substrate S (as shown in FIG. 1C), wherein the second suspended film F2 and the second cavity C2 can form a second resonance Structure. In the embodiment, the first and second piezoresistive elements P and P2 are respectively coupled to the first and second reflective electrodes R1 and R2 and respectively constitute a first loop L1 and a second loop L2. It is to be noted that the resonance frequencies of the first and second resonance structures may be determined by the size, material, and thickness of the substrate S, the suspended films FI and F2, and the cavities C1 and C2. As shown in FIG. 1A, since the resonant frequencies corresponding to the first and second resonant structures are different, when the acoustic waves of different frequencies in the external sound source W are received, the first and second suspended films F can be used. The first and second piezoresistive elements P and P2 on the F2 respectively sense the vibration of the different fundamental frequencies, and 9 200929170 can generate a corresponding equivalent resistive load in the first and second loops L1, L2. In operation, an AC signal can be input to the transmitting/receiving electrode E in the transmitting/receiving area e to generate an emission sound wave V (greater than 20K Hz) on the substrate B to the right first and second reflection areas r The first and second reflective electrodes R1 and R2 in r2; when the emitted acoustic wave V is transmitted to the first reflective region rl and the first resonant structural body can resonate with the first acoustic wave of the sound source W, the first reflective electrode R1 A first surface acoustic wave VI is generated according to the aforesaid emitted acoustic wave signal V and the first piezoresistive element P1 to the transmitting/receiving region e; the same is true; when the emitted acoustic wave V is transmitted to the second reflective region r2 and when the second resonant structural body The second reflective electrode R2 generates a second surface acoustic wave V2 to the transmitting/receiving area e according to the aforesaid emitted acoustic wave signal V and the second piezoresistive element P2, wherein the second reflective electrode R2 generates a resonance phenomenon with the second acoustic wave of the sound source W. The frequencies of the first and second sound waves are different. Referring again to FIG. 1D, the figure shows the time domain waveforms of the first and second surface acoustic waves VI and V2 return losses received by the transmitting/receiving electrode E, wherein the first and second reflective electrodes are The resistances of R1 and R2 are different, and the return loss values Ql and Q2 of the first and second surface acoustic waves VI and V2 are different, so that the first source of the sound source W can be known. , the frequency of the second sound wave and its intensity. As described above, the transmitting acoustic wave V can be transmitted to the first and second reflective regions rl, r2 through the transmitting/receiving region e, and then the first and second can be transmitted according to the first and second reflective electrodes R1 and R2. The surface acoustic waves VI and V2 respectively know the frequencies of the first and second acoustic waves in the sound source W and their corresponding intensity levels, wherein the transmitting/receiving electrode E can be connected to a control unit (not shown), 10 200929170 The unit may perform signal processing or related frequency domain analysis on the first and second surface acoustic waves VI, V2. Referring to FIG. 2A , a plurality of first and second suspended films FI and F2 and a plurality of first and second piezoresistive elements PI and P2 may be respectively disposed above the substrate S in another embodiment. The first piezoresistive elements P1 are respectively disposed on the first suspended film F1, and the second piezoresistive elements P2 are respectively disposed on the second suspended film F2. In particular, the first piezoresistive element P1 is coupled to the first reflective electrode R1 in parallel with each other, and the second piezoresistive element P2 is coupled to the second reflective electrode R2 in parallel with each other. As shown in FIGS. 2B to 2C, the first and second cavities C1 and C2 are formed under the first and second suspended films FI and F2, respectively, because the first and second suspended films FI and F2 are first and The second cavities C1 and C2 have different sizes and sizes, so that the first and second piezoresistive elements P1 and P2 on the first and second suspended films FI and F2 respectively sense the vibration of the upper phase and the fundamental frequency. . It should be particularly noted that the first and second suspended films FI and F2 and the first and second cavities C1 and C2 may form different resonant ❿ structures for different resonant frequencies, wherein the resonant frequency of the resonant structure It can be determined by parameters such as the size, material and thickness of the substrate S, the suspended films FI and F2, and the cavities C1 and C2. In this embodiment, by providing a plurality of parallel first piezoresistive elements P1 and second piezoresistive elements P2, not only the effect of multi-point reception can be achieved, but also the signal/noise ratio (S/N ratio) can be improved, and the part can be avoided. Structural damage leads to failure of the entire sound source analysis device, which in turn improves overall performance and stability. For example, the electrodes used in the conversion unit T and the suspended films F1 and F2 in the sound unit 200929170 can be formed on the substrate S through the microelectromechanical process to reduce the size and miniaturization of the system. The suspended film may also be replaced by an elongated suspension structure. Further, the substrate S may be a dielectric material such as (4) glass, and the transmitting/receiving electrode E and the first and second reflective electrodes Rb R2 are made of the material of the first (A1), and the substrate is composed of the electrical material of the county, for example. Aluminum nitride (A1N), zinc oxide (zn〇), lithium niobate (LiNb〇3), group acid clock __ (LiTa03) or lead zirconate titanate (PZT) may be used. Next, please refer to - and refer to f 3A, 3B 1]. In the first embodiment, the first and second piezoresistive members ΡΓ, P2' are first floating film F1'' disposed on the substrate s. The suspended film F1' is substantially circular and has a -th cavity C1' formed thereunder. It should be noted that the first and second piezoresistive 7G pieces PI', P2' > are not disposed at different positions on the first suspended film F1, so that the first suspended film F1 can be sensed separately. , the upper non-offset vibration 'and further through the aforementioned conversion unit T can obtain the extreme intensity of different sound waves in the sound source W. In summary, the present invention provides a sound source analyzing device, wherein the f-two sound-receiving unit and the conversion unit (for example, an interdigitated surface acoustic wave-changing pirate) can be used to detect differences in external sound sources. The sonic frequency and intensity 'not only the mechanism fl single can achieve the purpose of miniaturization. 12 200929170 [Simplified description of the drawing] Fig. 1A shows a schematic diagram of a sound source analyzing device; Fig. 1B shows a sectional view along Α1-ΑΓ in Fig. 1A; Fig. 1C shows a sectional view along A2-A2' in Fig. 1A 1D is a schematic diagram showing the time domain waveform of the return loss of the first and second surface acoustic waves; FIG. 2A is a schematic view showing another embodiment of the present invention; and FIG. 2B is a cross-sectional view taken along line B-B' of FIG. Fig. 2C is a cross-sectional view taken along line C-C' in Fig. 1A; Fig. 3A is a view showing another embodiment of the present invention; and Fig. 3B is a cross-sectional view taken along line D-D' in Fig. 3A. [Main component symbol description]
基板B 發射/接收區e 第一反射區rl ® 第二反射區r2 峰值Ql、Q2 發射聲波V 第一表面聲波VI 第二表面聲波V2 第一腔體C1、C1’Substrate B Transmitting/receiving area e First reflecting area rl ® Second reflecting area r2 Peak Ql, Q2 Acoustic wave V First surface acoustic wave VI Second surface acoustic wave V2 First cavity C1, C1'
第二腔體C2 發射/接收電極E 13 200929170 第一懸空薄膜FI、Fl’ 第二懸空薄膜F2 第一迴路L1 第二迴路L2 收音單元Μ 第一壓阻元件Ρ1、ΡΓ 第二壓阻元件Ρ2、Ρ2’ 第一反射電極R1 ❹ 第二反射電極R2Second cavity C2 Transmitting/receiving electrode E 13 200929170 First suspended film FI, Fl' Second suspended film F2 First loop L1 Second loop L2 Sound pickup unit Μ First piezoresistive element Ρ1, ΡΓ Second piezoresistive element Ρ2 , Ρ 2' first reflective electrode R1 ❹ second reflective electrode R2
基材S 轉換單元Τ 音源WSubstrate S conversion unit 音 source W
1414