201132135 post-process; and then to combine it with a microcontroller (MCU) by using flip-chip packaging technology, it will become a smart microphone array which receives sound from specific directions and contain the feature of high noisy removal. Especially suitable for hearing aids, high-efficiency radio microphone, voice identification, and sound source positioning. 四、指定代表圖: (一) 本案指定代表圖為:第(2 )圖。 (二) 本代表圖之元件符號簡單說明: 1.微麥克風陣列晶片 4.微控制器晶片 .〇·百几 2. 環狀凸塊 3. 金屬蓋 五、 本案若有化學式時,請揭示最能顯示發明特徵的化學式: 六、 發明說明: 【發明所屬之技術領域】 本發明係為-有關智慧型麥克風陣列的製作方法,特別是指 -種利用封裝技術而使其具有高度指向性的麥克風_。此智慧 ,麥克風_具有高指向性與絲雜訊的功能,特別適合於助聽 器、高收音功效的麥克風、語音辨識、音源定位等。 。 【先前技術】 在目前通訊n材㈣上,手機及辟麥克風•大主導市場的 201132135 關鍵,-般使时機上會遇到料不便之處,在吵_環境中無 法清楚聽騎方的聲音,而且自己所發㈣聲音相為環境秘 吵雜’使得財也無絲聽清楚,若钱游歧_可以大幅 提升傳統手齡克朗收音’以及過射錄音,這將會促使人 們在使用手機上便利許多。而且若能將麥克風_的聲音解析度 大幅财,並餘上後續演算法的處理,能達到高靈敏度的語:201132135 post-process; and then to combine it with a microcontroller (MCU) by using flip-chip packaging technology, it will become a smart microphone array which receives sound from specific directions and contain the feature of high noisy removal. Aids, high-efficiency radio microphone, voice identification, and sound source positioning. IV. Designated representative map: (1) The representative representative of the case is: (2). (2) Brief description of the symbol of the representative figure: 1. Micro-microphone array chip 4. Microcontroller chip. 〇·100. 2. Annular bump 3. Metal cover 5. If there is a chemical formula in this case, please reveal the most A chemical formula capable of displaying the characteristics of the invention: 6. Description of the Invention: [Technical Field] The present invention relates to a method for fabricating a smart microphone array, and more particularly to a microphone that is highly directional using a packaging technique _. This wisdom, microphone _ has high directivity and silk noise, especially suitable for hearing aids, high-frequency microphone, voice recognition, sound source positioning and so on. . [Prior technology] In the current communication n material (four), the mobile phone and the microphone • the leading market of 201132135 key, generally cause the inconvenience in the timing, in the noisy environment can not clearly listen to the sound of the rider And the sounds that I made (4) are the secrets of the environment, so that the money is not clear, if the money is ambiguous _ can greatly enhance the traditional hand-age krona and over-recording, which will prompt people to use the mobile phone. Convenient for many. Moreover, if the sound resolution of the microphone _ can be greatly increased, and the processing of the subsequent algorithm is left, a highly sensitive language can be achieved:
辨識,這些魏_就在於如何讓智慧.克風具有指向性的效 果以及增加麥克風的靈敏度。 傳統的指向性麥克風主要是__上的全向性麥克風利用 封裝方式組成方料列歧直_躲達到方向錄音效果,再 糟由電路上的配合制具有指向性的魏,所需要的面積與體積 較大。如果在單n例如面積2· 5mm*2.5mm,容納兩個以上全 向性麥克風,騎結_言聲音是如同-平面傳遞而來並非是所 熟知的球狀波,所以許多指向性麥克風都不做在晶片上,因為人 類月b所發出的聲音通常為丨GHz至lGKHz以内,假設在室溫下波長 的長度約為34m至0. 〇34m,若想要有良好的設計彼此間距d必須 要較遠,至少要大於波長的十六分之-,如圖1(a)所示,否則薄 膜文到的振動是同一種振幅,都為同步訊號,並非延遲訊號,無 法達到讓相位波形延遲的功能,所以彼此間距盡量越大越好,其 辨別性越高,因為麥克風間距若過度靠近,則對於聲音的辨識將 會一樣的訊號,則無法利用電路來分辨聲音來源,因此確實有需 要提出一種創新的設計與製造方法來縮小指向麥克風的體積與面 201132135 積。 夕目月’j國内外參考文獻,利用微機電製程所製作的微麥克風,大 夕採用體型微加或面型微加卫,此兩種加I技術幾乎成為所有 '政機電必備的技術。利用目前CM〇s原有的製程來剌_智慧型麥 ^風不單可崎低製作上的成本,而且在製造上品質穩定可大量 製& ’本發明提出-種創新的設計與製造方法結合微電鑄技術可 喊成共振腔及音㈣結構,以達有指向性麥克風的功能, Φ 這是傳統微機電麥克風所做不到的。 【發明内容】 本發明的目的是為了製作出具财指向性的智g型微麥克風 晶片,利用特有的CMOS製程實現在2.5_ χ 25麵的晶片大小 中疋成指向性麥克風晶片,其中指向性麥克風晶片設計原理是利 用兩種不同麥克風置於同—晶#上可以糾取到不同聲音的訊 _號’分別是主要聲源以及背景雜音,而兩者聲音無法影響到非對 應之麥克風,如圖1(b)所示,有些麥克風是收集到來自上方的聲 曰,有些則疋收集來自側面或是後方的聲音,可以達到讓相位波 形延遲的功能,將兩者聲音單獨收集交由微控制器進行後續的處 理。將上述的指向性麥克風晶片擷取到的訊號交由微控制器晶片 進行後續處理,搭配微控制器可將聲音進行後續處理,主要功用 在於纟'S強扎向性效果以及s吾音辨識功能,許多方法可以達到我們 所要求的,例如:獨立成分分析法(ICA)、相位延遲波束集中(Delay 201132135 and Sum Beamforming)、適應性濾波器等方式,以適應, 說明…健有_綠歧㈣紐料來 適應性濾波器(Adaptive filter)會將雜刪做調整,使得輪 =nO,故我們的輸出訊號:0utput = e =d(n)_vfn、侍剧出y 一 yu),如圖l(c) 所不,適應_波騎餘並非固定的,岐隨著輪人的參考^ 號來改變,其係數的改變量是由誤差項e來調整,若輪入訊號二 期望訊v號的統計特性是不穩定的時候,鱗由適應㈣波器所求Identification, these Wei _ is how to make wisdom. The wind has a directional effect and increase the sensitivity of the microphone. The traditional directional microphone is mainly the omnidirectional microphone on __, which uses the encapsulation method to form the arbitrarily arbitrarily arbitrarily _ hiding to achieve the direction recording effect, and then the directionality of the circuit on the circuit is directional, the required area and Larger size. If a single n, for example, an area of 2·5 mm*2.5 mm, accommodates more than two omnidirectional microphones, the riding sound is like a plane-transmitting ball that is not well known, so many directional microphones are not. On the wafer, because the sound emitted by human month b is usually within GHz to 1GKHz, assuming that the wavelength is about 34m to 0. 〇34m at room temperature, if you want to have a good design, the distance d must be Farther, at least greater than sixteenth of the wavelength - as shown in Figure 1 (a), otherwise the vibration of the film is the same amplitude, both synchronous signals, not delayed signals, can not reach the phase waveform delay Function, so the distance between each other should be as large as possible. The higher the discriminability, the more the microphone spacing is too close, the same signal will be recognized for the sound, and the circuit cannot be used to distinguish the sound source. Therefore, it is indeed necessary to propose an innovation. The design and manufacturing method is to reduce the volume of the pointing microphone with the face 201132135. Ximuyue's reference to domestic and foreign references, using micro-microphones made by MEMS process, the use of body-shaped micro-addition or face-type micro-enhancement, these two plus I technology has become almost all the necessary technology. Utilizing the current CM〇s original process 剌 _ 智慧 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦 麦Micro-electroforming technology can be called into the resonant cavity and sound (four) structure to achieve the function of a directional microphone, Φ which is not possible with traditional MEMS microphones. SUMMARY OF THE INVENTION The object of the present invention is to produce a wealthy g-type micro-microphone wafer, which realizes a directional microphone chip in a 2.5 χ 25-sided wafer size by using a unique CMOS process, wherein a directional microphone The principle of chip design is to use two different microphones to be placed on the same-crystal # to get different sounds. The signal 'number' is the main sound source and the background noise, and the two sounds cannot affect the non-corresponding microphone. As shown in 1(b), some microphones collect the sonar from above, while others collect the sound from the side or the rear to achieve the function of delaying the phase waveform. The two sounds are separately collected and passed to the microcontroller. Perform subsequent processing. The signal extracted by the above directional microphone chip is sent to the microcontroller chip for subsequent processing, and the sound can be subsequently processed by the microcontroller, and the main function is the 纟'S strong aligning effect and the sigma recognition function. Many methods can meet our requirements, such as: Independent Component Analysis (ICA), Phase Delay Beam Concentration (Delay 201132135 and Sum Beamforming), adaptive filters, etc., to adapt, explain... Jianyou _ Green (4) It is expected that the adaptive filter will adjust the miscellaneous deletion so that the round = nO, so our output signal: 0utput = e = d(n)_vfn, the waiter out y a yu), as shown in Figure l ( c) No, adaptation _ wave riding is not fixed, 岐 with the wheel's reference ^ number, the coefficient of change is adjusted by the error term e, if the round signal 2 expected signal v number statistics When the characteristic is unstable, the scale is obtained by the adaptive (four) wave device.
的解則必須不斷針職人減更新,如此方可得顺佳解。本發 明此設計具有的特色與優點如下: ⑴利用鮮互補錄半導體製減合覆晶縣技術可以完成麥 克風本身封裳’若覆晶的基板採用微控制器則可以達到系統 單一封裝化(system in packaging, SIP)的要求。 (2)智慧型麥克風陣列具有高指向性與去除雜訊的功能,而且單 體小’能廣泛應用在通訊產品上。 【實施方式】 本發明乃是一種智慧型微麥克風陣列製作方式。如圖2所示, 其智慧型微麥克風陣列主要係由微麥克風陣列晶片1、環狀凸塊 2、金屬蓋3、微控制器晶片4、印刷電路板5所結合而成。 微麥克風陣列晶片1係由全向性麥克風6,如圖3所示,以及 指向性麥克風7所構成,如圖4所示。全向性麥克風6結構如圖3 201132135 所示,其包含一具有侧孔8的振動薄膜9及具有若干通氣孔忉 之背板11,而在振動薄膜9與背板u之間有振動腔12。指向性 麥克風7結構如圖4所示,其包含—簡_ 9及具有若干通氣 孔ίο之背板11,而在振動薄膜9與背板u之間有振動腔12。微 麥克風陣列晶片1設計原理:全向性麥克風6想要接收聲音,必須 作用振動薄膜9上,而指向性麥克風7想要接收聲音,則必須作 用在振動_ 9上’設計全向性麥克風6收音方向*上方收音, 着而指向性麥克風7收音方向由下方收音,因此全向性麥克6和指 向性麥克風7所擷取到的聲音訊號會有所不同。 其製造方法可由CMOS後製程或MEMS製程所完成。其中 CMOS後製程製作方式如下: 步驟-:透過標準半導體製程,來設計微麥克風陣列晶片i佈局, 如圖5⑻所示,微麥克風陣列晶片i佈局,以〇. 35 um 製私為例,全向性麥克風6利用Metal-1層作為振動薄膜 φ 9, VIA-1 作為振動腔 12 ’ Metal-2,VIA-2,Metal-3, VIA-3,Metal-4等多層作為背板11。而指向性麥克風7 利用Metal-4層作為振動薄膜9, VIA-3作為振動腔12, Metal~l’ VIA-1 ’ Metal-2,VIA-2,Metal-3 等多層作為 背板11。 步驟二:將晶粒之正面旋塗一層光阻13後,進行曝光顯影定義出圖 案’再利用非等向性餘刻(如:RIE)將氧化;δ夕14钮刻完,如 圖5(b)所示。 201132135 步驟三:利用等向性侧溶液(如:Silox Vap〇x m)侧氧切μ,製 做出振動腔12,如圖5(c)所示。 步驟四:將晶粒之正面密合於她上,並將晶粒之背面旋塗一層光 阻13後’進行曝光顯影定義出圖案,再利用非等向性勉 刻(如:DRIE)將矽基材15蝕刻至想要的深度,完成微麥克 風陣列晶片1 ’如圖5(d)所示。 背板11的通氣孔10設置可以不與振動薄膜9上的孔8 相同位置,一般而言,振動薄膜9上的侧孔8的直徑要小於通 氣孔10但數量要多於通氣孔1G,例如通氣孔⑴的數量約為十個 以下。 MEMS製作方法如下: 步驟-:將錄材I5清洗乾淨,並於魏材之正_塗上—層光阻 13 ’進行曝光顯影之後,並且再沉積上一層導電薄膜16, 如圖6⑻所示。 步驟二將絲材之正面·上—層光阻13後,進行曝光顯影定義 出圖案,再沉積上-層絕緣層17,如圖6(b)所示。 步驟三:再將雜材之正面旋塗上—層光阻13後,進行曝光顯影定 義出圖案,再沉積上-層導電薄膜16,如圖6(c)所示。 步驟四··财基材之正面旋塗上—層光阻13後進行曝錢影定義 圖案並且在石夕基材之正面沉積上一層絕緣層Η,如圖 201132135 6⑷所示。 步驟五:將矽基材之正面旋塗上一声 ^ 13後’進行曝光顯影定義 ’、、^材之正面沉積上—層導電薄膜16,如圖 6(e)所示。 叫 步驟六:將矽基材之正面沉積絕緣層17,The solution must be constantly updated to the staff, so that you can get a good solution. The features and advantages of this design of the present invention are as follows: (1) The use of the fresh complementary recording semiconductor to reduce the crystal chip technology can complete the microphone itself. If the substrate is flipped, the microcontroller can achieve a single package system (system in Packaging, SIP) requirements. (2) The smart microphone array has high directivity and noise removal, and the single small unit can be widely used in communication products. [Embodiment] The present invention is a smart micro-microphone array production method. As shown in FIG. 2, the smart micro-microphone array is mainly composed of a micro-microphone array chip 1, an annular bump 2, a metal cover 3, a microcontroller chip 4, and a printed circuit board 5. The micro-microphone array wafer 1 is composed of an omnidirectional microphone 6, as shown in Fig. 3, and a directional microphone 7, as shown in Fig. 4. The structure of the omnidirectional microphone 6 is as shown in FIG. 3 201132135, which comprises a vibrating film 9 having side holes 8 and a back plate 11 having a plurality of vent holes, and a vibration chamber 12 between the vibrating film 9 and the back plate u. . The directional microphone 7 has a structure as shown in Fig. 4, which includes a _9 and a back plate 11 having a plurality of vent holes, and a vibration chamber 12 between the vibrating film 9 and the back plate u. Micro-microphone array chip 1 design principle: the omnidirectional microphone 6 wants to receive sound, must act on the vibrating film 9, and the directional microphone 7 wants to receive sound, it must act on the vibration _ 9 'design omnidirectional microphone 6 In the radio direction*, the radio is received above, and the directional microphone 7 receives the radio signal from below, so the omnidirectional microphone 6 and the directional microphone 7 will have different audio signals. The manufacturing method can be completed by a CMOS post process or a MEMS process. The CMOS post-process is as follows: Step-: Design the layout of the micro-microphone array chip through a standard semiconductor process, as shown in Figure 5 (8), the layout of the micro-microphone array chip, taking 〇. 35 um as an example, omnidirectional The microphone 6 uses the Metal-1 layer as the vibration film φ 9, VIA-1 as the vibration chamber 12 'Metal-2, VIA-2, Metal-3, VIA-3, Metal-4 and the like as the back sheet 11. The directional microphone 7 uses the Metal-4 layer as the vibrating film 9, and the VIA-3 as the vibrating chamber 12, the metal~l' VIA-1' Metal-2, VIA-2, Metal-3 and the like as the back sheet 11. Step 2: After the surface of the die is spin-coated with a layer of photoresist 13, the exposure and development are defined to define the pattern 're-use of anisotropic remnant (such as: RIE) to oxidize; the δ 夕 14 button is finished, as shown in Figure 5 ( b) shown. 201132135 Step 3: Using the isotropic side solution (eg Silox Vap〇x m) side oxygen cut μ, make the vibration chamber 12, as shown in Figure 5 (c). Step 4: The front side of the crystal grain is adhered to her, and the back surface of the crystal grain is spin-coated with a photoresist 13 to perform a development process to define a pattern, and then an anisotropic engraving (eg, DRIE) is used. The substrate 15 is etched to a desired depth to complete the micro-microphone array wafer 1' as shown in Figure 5(d). The venting opening 10 of the backing plate 11 may be disposed not at the same position as the hole 8 in the vibrating film 9. Generally, the side hole 8 on the vibrating film 9 has a diameter smaller than that of the vent hole 10 but more than the vent hole 1G, for example, The number of vent holes (1) is about ten or less. The MEMS fabrication method is as follows: Step-: The recording material I5 is cleaned, and after exposure and development is performed on the positive-coating-layer photoresist 13', and a conductive film 16 is deposited thereon, as shown in Fig. 6 (8). In the second step, after the front side of the wire and the upper layer photoresist 13, exposure and development are performed to define a pattern, and the upper-layer insulating layer 17 is deposited, as shown in Fig. 6(b). Step 3: After the front side of the miscellaneous material is spin-coated with the layer photoresist 13, the pattern is formed by exposure and development, and the upper-layer conductive film 16 is deposited, as shown in Fig. 6(c). Step 4: The front side of the financial substrate is spin-coated with a layer of photoresist 13 and then exposed to the shadow definition pattern and an insulating layer is deposited on the front side of the stone substrate, as shown in Fig. 201132135 6(4). Step 5: spin-coating the front side of the ruthenium substrate to perform the exposure development definition, and deposit the upper-layer conductive film 16 on the front side of the substrate, as shown in Fig. 6(e). Step 6: depositing an insulating layer 17 on the front side of the crucible substrate,
如圖6(ί)所示 6(g)所示 步驟八:將石夕基材之背面旋塗上一層光阻13後,利用非等向健刻 (如_财基材15_出想要的深度及圖案如圖啊 所示。 步驟八:將絲材之正面及麵旋塗上—層光阻叫,料背面進 行曝光顯影定義圖案,再使用等向⑽刻溶液(如:Η3ρ〇4) 將部分導電薄膜16去除,並且使用非等向性崎如:卿 去除部分絕緣層17,重覆此動作直刻至想要的深度, 如圖6(i)所示。 步驟九液用等向性餘刻(如:取⑹將部分導電_ 16去除,形成 振動腔12,完成微麥克風陣列晶片丨’如圖6(j)所示。 雖然上述的製程所揭示的麥克風結構,僅為一簡化的圖示, 然則該振動薄膜9可以是由至少-層金屬所構成,並具有殘餘應 力消除的機構,其上設有微小孔洞陣列作為餘刻孔8,該微小孔洞 201132135 的直徑應小於數微米,而且要小於背板的通氣孔1〇,而背板u 可以由多層金屬與介電層所疊成並具有若干通氣孔1〇以控制麥克 風阻尼。 微麥克風陣列晶片1之封裝方式至少有二: 方式-.如® 7所示’首先’訊號處理電路可以整合到麥克 風陣列晶片1上,也可以直接整合至另—微控制器⑼4上,藉 由印刷電路板5將上述兩晶片加以組裝,微控制器"4上的電 φ性輸出入接點與印刷電路板5的線路可以利用打線連接,並於印 w18 ’ _通微麥克鱗列⑼丄對應 的音腔,同時將金屬蓋3罩住兩晶片且其蓋底四週與印刷電路板5 使用黏著劑20密封接合。 方式二:如圖8所示,為了縮小封裝的面積,微麥克風陣列 晶片1可以覆晶疊合至另-微控制器晶片4上,覆晶疊合的方式, 包含訊號連接的獨立凸塊19,與振動腔的外圍成環狀凸塊2,該 籲環狀凸塊2可斷開成為音孔18,其結構如圖9所示。獨立凸塊二 與環狀凸塊2較佳是設立於微控繼晶片4上。藉由印刷電路板$ 將上述兩疊合晶片加以組裝,微控制器晶片4上的電性輸出入接 點與印刷電路板5的線路可以利用打線連接,同時將金屬蓋3罩 住兩晶片且其蓋底四週與印刷電路板5使用黏著劑2()密封接合, 並於金屬蓋3之側面部設置音孔18。請注意,上述的環狀凸塊2 形狀及孔洞18數量及大小可依實際運用進行設計。微麥克風陣列 晶片1經由覆晶封装接合後之等角視圖,如圖1〇所示。 201132135 —J、?:設:::::她片1可以傲在2.滅5 運碑=== 音,此知益丨貞异法運鼻過後’,说可以得到良好且清晰的聲 士 日慧型麥克風陣列可以應用在語音辨識以及藍芽耳機等通 机產品上。 [實施例一] 2χ2之智慧型麥克風陣列,如圖u (a)所示,可為一 #個全向性麥克風6和三個指向性麥克風7所組成,擺設 方式可為左上角為全向性麥克風6,在右上角、右下角、 左下角可擺设指向性麥克風7。或如圖η (b)所示,由兩 個全向性麥克風6和兩個指向性麥克風7所組成,左上 角、左下角可擺設全向性麥克風6,右上角、右下角可 擺設指向性麥克風7。或如圖11 (c)所示,可為兩個全向 性麥克風6和兩個指向性麥克風7所組成,擺設方式可 籲為右上角、右下角為全向性麥克風6,左上角、左下角 為指向性麥克風7。 [實施例二] 3 X 3之智慧塑麥克風陣列,如圖12⑷所示,其特 徵為由一個全向性麥克風6和八個指向性麥克風7所組 成’在陣列中心擺設全向性麥克風6,中心周圍八個方 向可擺設指向性麥克風7。或如圖12 (b)所示,其特徵 201132135 為五個全向性麥克6 陣列令心、左上角、右上“麥克風7所組成,在 向性麥角左下角、右下角可擺設全 =:二陣列中心的正上方、正下方、正右方、 工方了擺扠扣向性麥克風7。 徵為四個全向性麥克 / Θ (e)所7F ’其特 包太— 和五個指向性麥克風7所組 成’在母一邊的正中心掘# 、 左下角、右上角士 向性麥克風6,在左上角、 7。 角、右下角、正中心擺設置指向性麥克風 [實施例三] ==克風陣列’如圖13所示,其特徵係十字 ^ 王向性麥克風6 ’其餘皆為指向性麥克風7。 【圖式簡單說明】 圖1微控制器適應性據波器架構示意圖 •圖2智慧型麥克風陣列示意圖 圖3全向性麥克風結構示意圖 圖4指向性麥克風結構示意圖 圖5購-嶋製程製作微麥克風陣列流程示意圖 圖6 MEMS製程製作微麥克風陣列流程示意圖 圖7智慧型麥克風陣列晶片料方式—示意圖 圖8智慧型麥克風陣列晶片封裝方式二示意圖 201132135 圖9環狀凸塊示意圖 圖10微麥克風陣列立體圖As shown in Fig. 6(e), step 6 (g) shows the step 8: after the back surface of the stone substrate is spin-coated with a layer of photoresist 13, the non-isotropic carving is used. The depth and pattern are shown in the figure. Step 8: Apply the front and side of the wire to the layer-resistance, and define the pattern on the back of the material for exposure and development, and then use an isotropic (10) solution (eg: Η3ρ〇4). Part of the conductive film 16 is removed, and the non-isotropic is used to remove part of the insulating layer 17, and the action is repeated to the desired depth, as shown in Fig. 6(i). The directional residue (eg, taking (6) removes part of the conductive _ 16 to form the vibrating cavity 12, completes the micro-microphone array wafer 丨' as shown in FIG. 6(j). Although the microphone structure disclosed in the above process is only one In simplified illustration, the vibrating membrane 9 may be a mechanism composed of at least a layer of metal and having residual stress relief, and a micro-hole array is provided thereon as a residual hole 8, and the diameter of the micro-hole 201132135 should be less than the number Micron, and smaller than the vent hole of the back plate, and the back plate u can be composed of multiple layers of metal and dielectric layer It is stacked and has a number of vents 1 to control the microphone damping. The micro-microphone array chip 1 is packaged in at least two ways: mode - as shown in Figure 7, the 'first' signal processing circuit can be integrated into the microphone array chip 1, also It can be directly integrated into the other-microcontroller (9) 4, and the two wafers are assembled by the printed circuit board 5. The electric φ output on the microcontroller "4 and the printed circuit board 5 can be wired. Connecting, and printing the corresponding sound cavity of the w18 ' _ micro-micro-scale (9) ,, while covering the two wafers with the metal cover 3 and surrounding the bottom of the cover with the printed circuit board 5 using the adhesive 20 sealing joint. As shown in FIG. 8, in order to reduce the area of the package, the micro-microphone array wafer 1 can be flip-chip laminated onto the other-microcontroller chip 4, in a flip-chip manner, including the signal-connected independent bumps 19, and the vibration cavity. The outer periphery is formed into an annular protrusion 2, and the ring-shaped convex block 2 can be broken into a sound hole 18, and its structure is as shown in Fig. 9. The independent convex block 2 and the annular convex block 2 are preferably set up in the micro control. On wafer 4. By printed circuit board $ will The two stacked wafers are assembled, and the electrical input and output contacts on the microchip wafer 4 and the printed circuit board 5 can be connected by wire bonding, and the metal cover 3 covers the two wafers and the periphery of the cover and the printed circuit. The plate 5 is sealingly joined by an adhesive 2 (), and a sound hole 18 is provided on a side portion of the metal cover 3. Note that the shape and size and size of the annular projection 2 described above can be designed according to practical applications. The isometric view of the array wafer 1 after bonding through the flip chip package is as shown in Fig. 1. 201132135 — J, ?: Set::::: Her film 1 can be proud of 2. The 5th monument === tone, This know-how is different after the nose has passed, and it can be said that a good and clear voice can be applied to voice recognition and Bluetooth headsets. [Embodiment 1] The smart microphone array of 2χ2, as shown in Fig. u (a), may be composed of an omnidirectional microphone 6 and three directional microphones 7, and the arrangement may be omnidirectional to the upper left corner. The directional microphone 6 can be placed in the upper right corner, the lower right corner, and the lower left corner. Or as shown in Figure η (b), consisting of two omnidirectional microphones 6 and two directional microphones 7. The omnidirectional microphone 6 can be placed in the upper left corner and the lower left corner. The directivity can be set in the upper right corner and the lower right corner. Microphone 7. Or as shown in FIG. 11(c), it can be composed of two omnidirectional microphones 6 and two directional microphones 7. The arrangement can be called the upper right corner and the lower right corner is an omnidirectional microphone 6, upper left corner, lower left corner. The corner is a directional microphone 7. [Embodiment 2] The 3X3 smart plastic microphone array, as shown in Fig. 12(4), is characterized in that an omnidirectional microphone 6 and eight directional microphones 7 are composed of 'an omnidirectional microphone 6 disposed at the center of the array, A directional microphone 7 can be placed in eight directions around the center. Or as shown in Figure 12 (b), the feature 201132135 is composed of five omnidirectional microphone 6 arrays, the upper left corner, and the upper right "microphone 7, which can be placed in the lower left corner and the lower right corner of the erect erection =: The two sides of the array center are directly above, directly below, right side, and the work side has a forked directional microphone 7. Recruited as four omnidirectional microphones / Θ (e) 7F 'its special package too - and five points The sexual microphone 7 consists of 'the center of the mother's side of the dig #, the lower left corner, the upper right corner of the directional microphone 6, in the upper left corner, 7. The angle, the lower right corner, the center pendulum set the directional microphone [Embodiment 3] = = gram array] as shown in Figure 13, the characteristics of the cross ^ king of the microphone 6 'the rest are directional microphones 7. [Figure simple description] Figure 1 microcontroller adaptive data structure schematic diagram 2 Schematic diagram of smart microphone arrayFig.3 Schematic diagram of omnidirectional microphone structureFig.4 Schematic diagram of directional microphone structureFig.5 Schematic diagram of micro-microphone array flow diagram of purchase-嶋 process manufacturingFig.6 Schematic diagram of micro-microphone array process of MEMS processFig.7 Smart microphone array chip Material side --10 a schematic perspective view of the micro-array of microphones of the microphone array of FIG. 8 smart chip package of FIG. 9 201132135 embodiment two schematic block diagram of an annular projection
圖11 2x2智慧型麥克風陣列 示意圖 圖12 3 X 3智慧型麥克風陣列 示意圖 圖13十字智慧型麥克風陣列示意圖 【主要元件符號說明】 1.微麥克風陣列晶片 11.背板 2.環狀凸塊 12.振動腔 3.金屬蓋 13.光阻 4.微控制器晶片 14.氧化矽 5.印刷電路板 15.矽基材 6.全向性麥克風 16.導電薄膜 7.指向性麥克風 17.絕緣層 8.蝕刻孔 18.音孔 9.振動薄膜 19獨立凸塊 10.通氣孔 20黏著劑Fig. 11 Schematic diagram of 2x2 smart microphone array Fig. 12 Schematic diagram of 3 X 3 smart microphone array Fig. 13 Schematic diagram of cross smart microphone array [Description of main components] 1. Micro microphone array wafer 11. Back plate 2. Annular bump 12. Vibration chamber 3. Metal cover 13. Photoresist 4. Microcontroller wafer 14. Oxide oxide 5. Printed circuit board 15. Tantalum substrate 6. Omnidirectional microphone 16. Conductive film 7. Directional microphone 17. Insulation layer 8 Etched hole 18. Sound hole 9. Vibrating film 19 independent bump 10. Vent hole 20 adhesive