Small array micro-electromechanical microphone device and noise suppression method thereof

Abstract

A small array micro-electromechanical microphone device and a noise suppression method thereof. The device comprises: a mask; a circuit board coupled to the mask and comprising a first sound hole and a second sound hole; a first microphone module comprising: a first microphone and a first integrated circuit, wherein the first microphone extracts a first audio signal from a sound source through the first sound hole; and a second microphone module comprising: a second microphone and a second integrated circuit, wherein the second microphone extracts a second audio signal from the sound source through the second sound hole. The first microphone and the second microphone have the same sensitivity, phase and omnidirectionality. The first integrated circuit performs a first logic operation on the first and second audio signals to generate a first summed audio signal. The first integrated circuit performs a sampling delay on the second audio signal with a first clock signal, and subtracts the delayed second audio signal from the first audio signal to obtain a first differential audio signal with a first directivity. The second integrated circuit bypasses and outputs a second audio signal with omnidirectionality.

Description

Technical Field

[0001] The present invention relates to a microphone device, and in particular to a small array micro-electromechanical microphone device and a noise suppression method thereof Background Art

[0002] Most microphone devices today are condenser microphones, among which micro-electromechanical system (MEMS) microphones have been widely used. MEMS microphones use micro-electromechanical systems, which can integrate multiple functions of electronics, motors or mechanics in one device. Therefore, MEMS microphones have the advantages of small size, power saving, easy installation and anti-interference.

[0003] Generally speaking, microphone devices using multiple microphones (such as MEMS microphones) can have better sound sensitivity (sensitivity), signal-to-noise ratio (SNR) and directionality. However, if the microphone array uses omnidirectional microphones, the audio signal output by the microphone array often cannot have good directionality, so it is easy to provide an audio signal with more noise to the back-end voice recognition device, thereby greatly reducing the voice recognition rate.

[0004] Therefore, a small array micro-electromechanical microphone device and a noise suppression method thereof are needed to solve the above problems. Summary of the invention

[0005] The present invention provides a small array micro-electromechanical microphone device, comprising: a mask; a circuit board coupled to the mask and comprising a first sound hole and a second sound hole; a first microphone module, comprising: a first microphone and a first integrated circuit, wherein the first microphone extracts a first audio signal from a sound source through the first sound hole; and a second microphone module, comprising: a second microphone and a second integrated circuit, wherein the second microphone extracts a second audio signal from the sound source through the second sound hole, wherein the first microphone and the second microphone have the same sensitivity, phase and omnidirectionality, and the second audio signal is input to the first integrated circuit; wherein the first integrated circuit is used to perform a first logic operation on the first audio signal and the second audio signal to generate a first summed audio signal, wherein the first integrated circuit performs sampling delay on the second audio signal with a first clock signal, and subtracts the delayed second audio signal from the first audio signal to obtain a first differential audio signal, and the first differential audio signal has a first directivity, wherein the second integrated circuit bypasses and outputs the second audio signal, and the second audio signal has omnidirectionality.

[0006] In some embodiments, the first integrated circuit and the second integrated circuit are application-oriented integrated circuits, and the frequency of the first clock signal is 2.048 MHz. In addition, the first distance between the first microphone and the second microphone is 5 mm.

[0007] In some embodiments, the first integrated circuit further applies a low-pass filter and a first equalizer, a band-pass filter and a second equalizer, and a high-pass filter and a third equalizer to the first summed audio signal, the first differential audio signal, and the second audio signal to generate a first low-frequency audio signal, a first intermediate-frequency audio signal, and a first high-frequency audio signal, and the sensitivities of the first low-frequency audio signal, the first intermediate-frequency audio signal, and the first high-frequency audio signal are substantially the same. The pass band of the low-pass filter is below 200 Hz, the pass band of the band-pass filter is 200 to 7400 Hz, and the pass band of the high-pass filter is above 7400 Hz.

[0008] In some embodiments, the first audio signal is further input to the second integrated circuit, and the second integrated circuit performs sampling delay on the first audio signal with the first clock signal, and subtracts the delayed first audio signal from the second audio signal to obtain a second differential audio signal, and the second differential audio signal has a second directivity different from the first directivity. The second integrated circuit further applies the low-pass filter and the first equalizer, the band-pass filter and the second equalizer, and the high-pass filter and the third equalizer to the second summed audio signal, the second differential audio signal, and the first audio signal to generate a second low-frequency audio signal, a second intermediate-frequency audio signal, and a second high-frequency audio signal. The second integrated circuit also sums the second low-frequency audio signal, the second intermediate-frequency audio signal, and the second high-frequency audio signal to generate a second beam, and the second beam has the second directivity.

[0009] The present invention further provides a noise suppression method for a small array micro-electromechanical microphone device, the small array micro-electromechanical microphone device comprising a mask; a circuit board; a first microphone module comprising: a first microphone and a first integrated circuit; and a second microphone module comprising: a second microphone and a second integrated circuit, wherein the first microphone and the second microphone have the same sensitivity, phase and omnidirectionality, the method comprising: using the first microphone and the second microphone to extract a first audio signal and a second audio signal from a sound source respectively; using the first integrated circuit to perform a first logic operation on the first audio signal and the second audio signal to generate a first summed audio signal; using the first integrated circuit to sample and delay the second audio signal with a first clock signal, and subtracting the delayed second audio signal from the first audio signal to obtain a first differential audio signal, wherein the first differential audio signal has a first directivity; and using the second integrated circuit to bypass and output the second audio signal, wherein the second audio signal has omnidirectionality.

[0010] The present invention further provides a small array micro-electromechanical microphone device, comprising: a mask; a circuit board coupled to the mask and comprising a first sound hole, a second sound hole and a third sound hole; a first microphone module, comprising: a first microphone and a first integrated circuit, wherein the first microphone extracts a first audio signal from a sound source through the first sound hole; and a second microphone module, comprising: a second microphone and a second integrated circuit, wherein the second microphone extracts a second audio signal from the sound source through the second sound hole; a third microphone module, comprising: a third microphone and a third integrated circuit, wherein the third microphone extracts a third audio signal from the sound source through the third sound hole, wherein the first microphone, the second microphone and the third integrated circuit are connected to the first microphone module and the second microphone module. The microphone and the third microphone have the same sensitivity, phase and omnidirectionality, wherein the first audio signal is input to the second integrated circuit, the second audio signal is input to the third integrated circuit, and the third audio signal is input to the first integrated circuit; wherein the first integrated circuit is used to perform a first logic operation on the first audio signal and the third audio signal to generate a first summed audio signal, and to perform sampling delay on the third audio signal with a first clock signal, and to subtract the delayed third audio signal from the first audio signal to obtain a first differential audio signal, and the first differential audio signal has a first directivity, wherein the first integrated circuit bypasses and outputs the first audio signal, and the first audio signal has omnidirectionality. The second integrated circuit is used to perform a second logic operation on the second audio signal and the first audio signal to generate a second summed audio signal, and to perform sampling delay on the first audio signal with the first clock signal, and to subtract the delayed first audio signal from the second audio signal to obtain a second differential audio signal, and the second differential audio signal has a second directivity, wherein the second integrated circuit bypasses and outputs the second audio signal, and the second audio signal has omnidirectionality. The third integrated circuit is used to perform a third logic operation on the third audio signal and the second audio signal to generate a third summed audio signal, and to perform sampling delay on the second audio signal using the first clock signal, and to subtract the delayed second audio signal from the third audio signal to obtain a third differential audio signal, and the third differential audio signal has a third directivity, wherein the third integrated circuit bypasses and outputs the third audio signal, and the third audio signal has omnidirectionality. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1A FIG. 4 is a schematic diagram of a small array MEMS microphone device according to an embodiment of the present invention.

[0012] Figure 1B According to the present invention Figure 1A A perspective view of a small array MEMS microphone device according to an embodiment.

[0013] Figure 1CFIG. 4 is a perspective view of a small array MEMS microphone device according to another embodiment of the present invention.

[0014] Figure 1D FIG. 4 is a perspective view of a small array MEMS microphone device according to another embodiment of the present invention.

[0015] Figure 1E FIG. 4 is a perspective view of a small array MEMS microphone device according to another embodiment of the present invention.

[0016] Figure 2 FIG. 4 is a block diagram of a small array MEMS microphone device according to an embodiment of the present invention.

[0017] Figure 3A is a block diagram of a microphone module according to an embodiment of the present invention.

[0018] Figure 3B FIG. 4 is a schematic diagram of a microphone module in combination with a processor to flatten audio response according to an embodiment of the present invention.

[0019] Figure 4A FIG. 4 is a schematic diagram of a frequency response curve of a differential audio signal according to an embodiment of the present invention.

[0020] Figure 4B FIG. 4 is a schematic diagram of a frequency response curve of an output audio signal according to an embodiment of the present invention.

[0021] Figure 5A Schematic diagram of a microphone array using two omnidirectional microphones.

[0022] Figure 5B Polar diagram of a traditional microphone array.

[0023] Figure 6 is a block diagram of a microphone module according to another embodiment of the present invention.

[0024] Figure 7 According to the present invention Figure 6 Block diagram of a microphone module in an embodiment.

[0025] Figure 8 According to the present invention Figure 6 A schematic diagram of the connection method of the microphone module in the embodiment.

[0026] Fig. 9 According to the present invention Figure 8 A schematic diagram of the connection method of the microphone module in the embodiment.

[0027] Fig.10 According to the present invention Figure 8 A schematic diagram of the connection method of the microphone module in the embodiment.

[0028] Fig.11 FIG. 4 is a flow chart of a noise suppression method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The following description lists various embodiments of the present invention. The following description introduces the basic concept of the present invention and is not intended to limit the content of the present invention. The actual scope of the invention should be defined in accordance with the claims.

[0030] Figure 1A FIG. 4 is a schematic diagram of a small array MEMS microphone device according to an embodiment of the present invention. Figure 1B According to the present invention Figure 1A A three-dimensional diagram of a small array micro-electromechanical microphone device according to an embodiment. Please also refer to Figure 1A and Figure 1B The small array micro-electromechanical microphone device 100 includes a mask 101, a circuit board 102, a processor 103, a microphone module 110 and a microphone module 120. The processor 103 couples the mask 101 and the circuit board 102 to form a chamber CH1 and a chamber CH2. The microphone module 110 in the chamber CH1 includes a diaphragm 111, a microphone 112 and an integrated circuit 113. The microphone module 120 in the chamber CH2 includes a diaphragm 121, a microphone 122 and an integrated circuit 123. The circuit board 102 is coupled to the mask 101 and includes a sound hole 104 and a sound hole 105, and the sound hole 104 and the sound hole 105 are separated by a distance d0 (for example, 5 mm, not limited). The sound holes 104 and 105 (for example, which can be regarded as front / rear sound holes) in the small array micro-electromechanical microphone device 100 are located on the same plane and receive audio signals through the diaphragms 111 and 121 respectively. In some embodiments, the length d0 is the distance between the center point of the sound hole 104 and the center point of the sound hole 105 , and may also be regarded as the distance between the microphone 112 and the microphone 122 .

[0031] In some embodiments, the processor 103 may be a digital signal processor or a microcontroller. The microphones 112 and 122 are micro-electromechanical system (MEMS) devices and may form a microphone array. In some embodiments, the integrated circuits 113 and 123 may be, for example, an application-oriented integrated circuit, which includes a digital circuit (e.g., a circuit that can perform digital signal processing), an analog circuit (e.g., an operational amplifier), an analog / digital conversion circuit, and an operational circuit, the details of which will be described in detail below.

[0032] In some embodiments, the integrated circuits 113 and 123 can be connected to and control the microphones 112 and 122, respectively. In some embodiments, the integrated circuits 113 and 123 can be connected to the circuit board 102 through conductors (or wires), and then coupled to the microphones 112 and 122 through other conductors, thereby providing voltage to the microphones 112 and 122 and receiving and processing the signals generated by the microphones 112 and 122 based on the sound. In some embodiments, the integrated circuits 113 and 123 can use a pulse-density modulation (PDM) protocol on the circuit board 102 for data transmission.

[0033] In some embodiments, the material of the mask 101 is metal, and a groove VP is formed on the mask 101. On the other hand, when the material of the mask 101 is metal, the thickness of the mask 101 can be reduced and still have sufficient rigidity, thereby further reducing the volume of the small array MEMS microphone device 100.

[0034] In some embodiments, the volume of the chamber CH1 is the same as that of the chamber CH2, and the configuration of the microphone module 110 in the chamber CH1 is the same as that of the microphone module 120 in the chamber CH2. In this case, the environment corresponding to the microphone module 110 is substantially the same as that corresponding to the microphone module 120. Therefore, when the integrated circuits 113 and 123 process the signals generated by the microphones 112 and 122 and perform the directivity function of the small array micro-electro-mechanical microphone device 100, the influence caused by the environmental error between the microphones 112 and 122 can be reduced, thereby improving the accuracy of the directivity of the small array micro-electro-mechanical microphone device 100.

[0035] In some embodiments, the integrated circuits 113 and 123 can provide the same voltage to the microphones 112 and 122, so that the distance between the diaphragm 111 of the microphone 112 and the back plate (not shown) of the microphone 112 is the same as the distance between the diaphragm 121 of the microphone 122 and the back plate (not shown) of the microphone 122, thereby allowing the microphones 112 and 122 to have the same sound sensitivity, thereby allowing the small array micro-electromechanical microphone device 100 to have a better signal-to-noise ratio.

[0036] like Figure 1AAs shown, the position of the sound hole 104 corresponds to the position of the diaphragm 111, so that the diaphragm 111 can receive the sound signal through the sound hole 104; and the position of the sound hole 105 corresponds to the position of the diaphragm 121, so that the diaphragm 121 can receive the sound signal through the sound hole 105. In some embodiments, a first sound signal outside the small array micro-electromechanical microphone device 100 can be transmitted to the microphone 110 and the microphone 120 through the sound holes 104 and 105 respectively. Since the sound holes 104 and 105 are separated by a length d0 (e.g., 5 mm), when the first sound signal has a specific transmission direction, the time when the first part of the sound signal is transmitted to the diaphragm 111 can be the same as the time when the second part of the sound signal is transmitted to the diaphragm 121, thereby making the small array micro-electromechanical microphone device 100 have directivity. In some embodiments, the length d0 is the distance between the center point of the sound hole 104 and the center point of the sound hole 105, and can also be regarded as the distance between the microphone 110 and the microphone 120.

[0037] In some embodiments, the sound signal transmitted from the sound hole 104 to the diaphragm 111 (e.g., the first part of the sound signal of the first sound signal) will not be transmitted to the diaphragm 121, and the sound signal transmitted from the sound hole 105 to the diaphragm 121 (e.g., the second part of the sound signal of the first sound signal) will not be transmitted to the diaphragm 111. In this case, the sound signal transmitted to the microphone 120 in the chamber CH2 will not interfere with the microphone 110 in the chamber CH1; similarly, the sound signal transmitted to the microphone 110 in the chamber CH1 will not interfere with the microphone 120 in the chamber CH2. Therefore, the noises received by the microphones 110 and 120 can be reduced, thereby improving the directivity performance of the microphone device 100.

[0038] Figure 1C FIG. 4 is a perspective view of a small array MEMS microphone device according to another embodiment of the present invention.

[0039] like Figure 1C As shown, in another embodiment, the microphone modules 110 and 120 are disposed on the circuit board 102 , but have separate housings 101A and 101B, respectively. In addition, the processor 103 is disposed in the microphone module 120 and is stacked with the integrated circuit 123 .

[0040] Figure 1D FIG. 4 is a perspective view of a small array MEMS microphone device according to another embodiment of the present invention.

[0041] like Figure 1DAs shown, in another embodiment, the microphone modules 110 and 120 are independent modules, and have independent circuit boards 102A and 102B, and independent housings 101A and 101B, respectively. In addition, the processor 103 is disposed in the microphone module 110 and is overlapped with the integrated circuit 113. When a small microphone array is to be implemented, the distance d between the microphones 112 and 122 in the microphone modules 110 and 120 is, for example, 5 mm. Because the microphone modules 110 and 120 can be disposed separately, if a microphone array with a longer microphone distance is required in the electronic device, the distance d between the microphones 112 and 122 can be, for example, 20 mm. In this embodiment, if the electronic device requires a microphone array formed by three or more microphone modules, one microphone module 110 (including the processor 103) and two or more microphone modules 120 can be disposed on the circuit board of the electronic device.

[0042] Figure 1E FIG. 4 is a perspective view of a small array MEMS microphone device according to another embodiment of the present invention.

[0043] like Figure 1E As shown in FIG. 1 , in another embodiment, the microphone modules 110 and 120 are independent modules and have independent circuit boards 102A and 102B, and independent housings 101A and 101B, respectively. Figure 1D The embodiment is different in that the processor 103 is separately arranged outside the microphone modules 110 and 120. Similarly, when a small microphone array is to be implemented, the distance d between the microphones 112 and 122 in the microphone modules 110 and 120 is, for example, 5 mm. Because the microphone modules 110 and 120 can be separately arranged, if a microphone array with a longer microphone distance is required in the electronic device, the distance d between the microphones 112 and 122 can be, for example, 20 mm. In this embodiment, the microphone modules 110 and 120 are substantially the same microphone modules. If the electronic device requires a microphone array formed by 3 or more microphone modules, three microphone modules 110 can be used and then combined with the processor 103 to be arranged on the circuit board of the electronic device, and the distance between any two microphone modules can be set according to actual needs.

[0044] Figure 2 FIG. 4 is a block diagram of a small array MEMS microphone device according to an embodiment of the present invention.

[0045] Please also refer to Figure 1A and Figure 2. Microphone modules 110 and 120 both include CS, DATA, CLK, PDM_in, L / R and other pins. Processor 103 can transmit control signal 201 to CS pin (chip select) of microphone modules 110 and 120 to control microphone modules 110 and 120 to turn on or off. Processor 103 also provides clock signal 202 to CLK pin (clock) of microphone modules 110 and 120 to provide clock signal (e.g., 2.048 MHz clock signal) required for integrated circuits 113 and 123 in microphone modules 110 and 120 to operate. L / R pin of microphone modules 110 and 120 defines whether microphones 112 and 122 in microphone modules 110 and 120 are used for front microphones or rear microphones. In this embodiment, L / R pin of microphone module 110 is grounded, indicating that microphone 112 is a front microphone. L / R pin of microphone module 120 is connected to voltage VDD, indicating that microphone 122 is a rear microphone.

[0046] The PDM_in pins of the microphone modules 110 and 120 are input signal pins, and the input signals can be subjected to logic operations and time delay processing with the audio signals extracted by the microphones 112 and 122, and the audio signals 203 and 204 are respectively output from the DATA pins of the microphone modules 110 and 120 to the DATA_IN pin of the processor 103. In this embodiment, due to design requirements, the PDM_in pin of the microphone module 120 does not input any audio signal, and the audio signal 204 output by the microphone module 120 is, for example, the audio signal (MicB) extracted by the microphone 122 and bypassed to the DATA pin to output the audio signal 204. The audio signal 204 is input to the PDM_in pin of the microphone module 110, and the audio signal 203 output by the microphone module 110 may include a sum signal (SUM) and a channel signal (DMA), wherein the sum signal SUM is, for example, obtained by adding the audio signal MicF extracted by the microphone 112 and the audio signal 204 (i.e., the audio signal MicB) input to the PDM_in pin. The channel signal DMA is obtained by subtracting the product of the time delay τ and the audio signal MicB from the audio signal MicF, i.e., DMA=MicF-MicB×τ.

[0047] Figure 3A is a block diagram of a microphone module according to an embodiment of the present invention.

[0048] Please also refer to Figure 2 and Figure 3A The microphone modules 110 and 120 are Figure 2 For ease of explanation, Figure 3AThe embodiment of the present invention takes the microphone module 110 as an example, and the configuration of the microphone module 120 is similar to the microphone module 110, the difference being that the PDM_in pin of the microphone module 120 has no input signal.

[0049] The integrated circuit 113 of the microphone module 110 includes a preamplifier 311 , an analog-to-digital converter (ADC) 312 , an adder 315 , an output control circuit 316 , a multiplexer 317 , a delay circuit 318 , a control circuit 319 , a subtractor 320 , and an equalizer 321 .

[0050] For example, the audio signal 301 (e.g., an analog signal) picked up by the microphone 112 needs to be amplified by the preamplifier 311 to generate an audio signal 341 (e.g., an analog signal) for subsequent processing, wherein the audio signal 341 can be, for example, referred to as a MicF audio signal. The PDM_in pin 325 of the integrated circuit 113 receives the MicB audio signal output by the integrated circuit 123, for example, an audio signal 345 encoded using the PDM protocol.

[0051] The adder 315 adds the audio signal 342 (MicF audio signal) and the audio signal 345 (MicB audio signal) to obtain an audio signal 343 (summed audio signal SUM). In addition, the MicB audio signal passes through the multiplexer 317 and the delay circuit 318 to generate a delayed audio signal 344, and the subtractor 320 subtracts the audio signal 345 from the audio signal 342 to obtain a differential audio signal 346. The audio signal 346 is then processed by the equalizer 321 to obtain an audio signal 347, wherein the audio signal 347 may be referred to as a differential audio signal DMA, for example. The output control circuit 316 may receive the audio signals 343 and 347, and select the audio signal 343 and / or the audio signal 347 to be output to generate corresponding output audio signals 331 (summed audio signal SUM) and 332 (differential audio signal DMA). It should be noted that the PDM_in pin of the microphone module 120 has no input signal, so the microphone module 120 may be output via a circuit similar to Figure 3A to output the MicB audio signal.

[0052] Figure 3B FIG. 4 is a schematic diagram of a microphone module in combination with a processor to flatten audio response according to an embodiment of the present invention.

[0053] Please also refer to Figure 2 , Figure 3A and Figure 3BThe hardware part 30 includes microphone modules 110 and 120, and the components in the software part 40 are executed by the processor 103. The microphone module 110 can output audio signals 331 (summed audio signal SUM) and 332 (differential audio signal DMA) to the processor 103, and the microphone module 110 can output MicB audio signal 333 to the processor 103. The processor 103 applies different filters to different input audio signals SUM, DMA and MicB.

[0054] For example, the processor 103 applies a low-pass filter 370 to the summed audio signal SUM, applies a band-pass filter 372 to the differential audio signal DMA, and applies a high-pass filter 374 to the MicB audio signal, wherein the passing frequency of the low-pass filter 370 is approximately below 200 Hz (or 20 to 200 Hz), the passing frequency of the band-pass filter 372 is approximately 200 to 7400 Hz, and the passing frequency of the high-pass filter 374 is approximately above 7400 Hz.

[0055] like Figure 4A As shown, curves 402, 404, 406 and 408 represent the frequency response curves of the differential audio signal DMA332 when the microphone distance d0 is 5 mm, 10 mm, 15 mm and 20 mm, respectively. It can be seen that the frequency response curve of the differential audio signal DMA332 is not flat, and the lower the frequency of the differential audio signal DMA332, the lower its gain value. In addition, when the microphone distance d0 is larger, the cutoff frequency of the corresponding frequency response curve is lower, such as Figure 4A As shown, the cut-off frequency wc_20 of the curve 408 is less than the cut-off frequencies wc_15, wc_10 and wc_5 (not shown) of the other curves 406, 404 and 402. The differential audio signal DMA332 passes through the bandpass filter 372 and then is processed by the equalizer 382 to obtain the intermediate frequency audio signal 392, and the gain value of the frequency response curve of the intermediate frequency audio signal 392 at the intermediate frequency (200-7400 Hz) can be flattened (substantially equalized). Similarly, after the summed audio signal SUM331 passes through the low-pass filter 370, it will be processed by the equalizer 380 to obtain the low-frequency audio signal 391. After the MicB audio signal 333 passes through the high-pass filter 374, it will be processed by the equalizer 384 to obtain the high-frequency audio signal 393.

[0056] Finally, the adder 386 adds the low-frequency audio signal 391, the intermediate-frequency audio signal 392, and the high-frequency audio signal 393 to obtain an output audio signal 390, wherein the frequency response curve of the audio signal 390 is as follows: Figure 4BAs shown, the gain values ​​at low frequencies (less than 200 Hz), medium frequencies (200-7400 Hz) and high frequencies (above 7400 Hz) can be flattened (substantially balanced).

[0057] Figure 5A Schematic diagram of a microphone array using two omnidirectional microphones. Figure 5B Polar diagram of a traditional microphone array.

[0058] A conventional microphone array for noise suppression can be implemented, for example, using two pre-set microphones, such as a combination of a unidirectional microphone (abbreviated as U-Mic) and an omnidirectional microphone (abbreviated as O-Mic) (e.g., a first combination), or using two omnidirectional microphones (e.g., a second combination).

[0059] Taking the first combination as an example, the traditional directional electret microphone mainly uses the difference in the paths of the sound from the front sound hole and the rear sound hole of the microphone to generate the difference in sound pressure of the two paths, and uses this difference in sound pressure to generate the displacement of the diaphragm to generate the difference in capacitance of the microphone, so that a signal output can be formed to achieve the fixed directivity of the microphone array. If a microphone array composed of a directional microphone and an omnidirectional microphone is to be used in an electronic device, a directional microphone with a front sound hole and a rear sound hole needs to be set, so a front / rear sound guide tube needs to be designed in the mechanism, which will increase the difficulty of designing the electronic device. If the electronic device does not allow the opening design on the rear appearance, the directivity of the microphone array will be lost. If the directivity of the microphone array is to be maintained, the design of the sound guide tube introduced into the rear sound hole of the microphone must be introduced from the front of the electronic device. However, the different lengths of the front / rear sound guide tubes will cause the directivity of the original simple microphone array to be different from the directivity of the microphone array after the microphone array is placed in the mechanism of the electronic device, thereby causing inconvenience in use and reduced noise suppression efficiency.

[0060] Taking the second combination as an example, please refer to Figure 5A In the second combination of microphone arrays, the audio signals detected by the two microphones are generated by the vibration of the two microphone diaphragms, and whether the characteristics of the two microphone diaphragms are consistent will affect the consistency of the audio signals of the two. However, it is difficult to achieve consistency in the film stretching and corona process of the microphone diaphragm of the traditional electret microphone. If both microphones 511 and 512 of the microphone array 500 use micro-electromechanical microphones, the consistency of the diaphragms of the two microphones 511 and 512 can be improved.

[0061] If the microphone array 500 is implemented by two micro-electromechanical (MEMS) omnidirectional microphones 511 and 512, when the sound source 520 emits sound, the audio signals XF and XB detected by the microphone 511 (front microphone) and the microphone 512 (rear microphone) are respectively summed and delayed to obtain a differential signal Pd that has directivity and can achieve the function of suppressing noise, wherein the differential signal P d =X F -X B ×τ, where delay τ=d0 / c, d0 is the distance between microphones 511 and 512, and c is the speed of sound. In addition, it should be noted that if a small microphone array consisting of two omnidirectional microphones (distance d0 is about 5 mm) is to generate a beam, the accuracy of the sample delay τ is also required.

[0062] Figure 5B Curves 508, 506, 504 and 502 are polar patterns obtained when the distance between microphones 511 and 512 is 5 mm and the software sampling rates are 16 KHz, 32 KHz, 48 KHz and 2.048 MHz respectively. If the sampling delay is generated by the software (i.e., the processor), the accuracy of the sampling delay is determined by the software sampling rate. However, if the software sampling rate is low, the originally required beam pointing will be changed from a single directional characteristic (e.g., a cardioid polar pattern, such as Figure 5B 502) and becomes a dual-directional characteristic (for example, a dipole polarization pattern, such as Figure 5B Although increasing the sampling rate of the software can improve the accuracy of the sampling delay, the software operation time will be longer and the power consumption of the processor will also increase.

[0063] If the distance d0 between microphones 511 and 512 is increased, the sampling delay value can be increased, and a lower software sampling rate can be used to achieve the required sampling delay accuracy. However, the increased microphone distance d0 cannot meet the requirements of the small microphone array, and the cutoff frequency wc of the frequency response of the formed beam will also be reduced. Therefore, the effective bandwidth of the microphone array becomes narrower, which in turn affects the ability of the small array micro-electromechanical microphone device to effectively eliminate noise, and the output audio signal will greatly reduce the speech recognition rate of the back-end speech recognition device.

[0064] Figure 6 is a block diagram of a microphone module according to another embodiment of the present invention.

[0065] Please also refer to Figure 3A-3B and Figure 6 , the difference is Figure 3BThe software part 40 in Figure 6 All are implemented by hardware circuits, that is, represented as hardware part 60, wherein the hardware part 60 may be included in the integrated circuits 113 and 123, or implemented by other digital circuits in the small array micro-electromechanical microphone device 100. The hardware circuits of the microphone modules 110 and 120 that generate the summed audio signal SUM, the differential audio signal DMA and the MicB audio signal can be referred to. Figure 3A , so the details are not repeated here. For example, when the bypass mode is not turned on, switches 601 and 602 are in an open state, and the summed audio signal SUM generated by the microphone module 110 can be input to the low-pass filter 610 to generate an audio signal 632, the differential audio signal DMA can pass through the high-pass filter 612 and the low-pass filter 614 (for example, both can be regarded as band-pass filters, and their order can be interchanged) to generate an audio signal 634, and the MicB audio signal can pass through the high-pass filter 616 to generate an audio signal 636. The audio signals 632, 634 and 636 are then summed up by the adding circuit 618 to generate an audio signal 638. The audio signal 638 is similar to Figure 3B The frequency response curve of the audio signal 638 is also as follows: Figure 4B As shown, the gain values ​​at low frequencies (less than 200 Hz), medium frequencies (200 to 7400 Hz), and high frequencies (above 7400 Hz) can be flattened (substantially balanced). In some embodiments, the audio signals 632, 634, and 636 can be processed similar to Figure 3B The signals are processed by the equalizers 380 to 384 and then input to the adding circuit 618.

[0066] In addition, when the bypass mode is turned on, switches 601 and 602 are both closed, and the summed audio signal SUM and the MicB audio signal are respectively transmitted to the multiplexer 620, and the multiplexer 620 can select to output audio signals 641 and / or 642 according to the control signal, wherein the audio signal 641 represents a directional beam, or can be called a U-Mic audio signal, and the audio signal 642 represents an omnidirectional audio signal, or can be called an O-Mic audio signal.

[0067] Figure 7 According to the present invention Figure 6 Block diagram of a microphone module in an embodiment.

[0068] exist Figure 6 In the embodiment of the present invention, each microphone module can output a summed audio signal SUM or a bypass (BYPASS) audio signal, and a differential audio signal DMA, and the integrated circuit in each microphone module includes functions such as an adder, a subtractor, DMA (differential delay), and an equalizer (EQ). Figure 7The connection method of the microphone modules 110 and 120 shown uses the MicF audio signal output by the microphone module 110 as the input signal of the PDM_in pin of the microphone module 120, and uses the MicB audio signal output by the microphone module 120 as the input signal of the PDM_in pin of the microphone module 110. Therefore, the microphone modules 110 and 120 can provide two directional audio signals (beams Beam1 and Beam2) with different polar patterns and directivities and omnidirectional audio signals (such as MicF and MicB audio signals) for use by the subsequent platform of the microphone module, that is, the microphone module can provide a first combination of small microphone arrays (U-Mic+O-Mic) to the subsequent platform to achieve better noise suppression effect.

[0069] Figure 8 According to the present invention Figure 6 A schematic diagram of the connection method of the microphone module in the embodiment.

[0070] exist Figure 8 In FIG. 1 , microphone module 130 is the same as microphone modules 110 and 120. Figure 7 In the embodiment of FIG. 1 , it can be known that through appropriate connection, the microphone modules 110 and 120 can output directional audio signals (beams Beam1 and Beam2) and omnidirectional audio signals (such as MicF and MicB audio signals) with different polar patterns and directions. Figure 8 In the embodiment, the microphone modules 110, 120 and 130 can also be connected in an appropriate manner so that the microphone modules 110, 120 and 130 can output directional audio signals (beams Beam1, Beam2 and Beam3) and omnidirectional audio signals (such as MicF, MicB and MicA audio signals) with different polar patterns and directivities.

[0071] For example, the MicF audio signal output by the microphone module 110 is input to the PDM_in pin of the microphone module 120, the MicB audio signal output by the microphone module 120 is input to the PDM_in pin of the microphone module 130, and the MicA audio signal output by the microphone module 130 is input to the PDM_in pin of the microphone module 110, which means that the integrated circuit 113 in the microphone module 110 processes the MicF audio signal and the MicA audio signal, the integrated circuit 123 in the microphone module 120 processes the MicB audio signal and the MicF audio signal, and the integrated circuit in the microphone module 130 processes the MicA audio signal and the MicB audio signal. Therefore, the integrated circuit in each microphone module operates on the two input audio signals to obtain a directional differential audio signal (beam) and an omnidirectional audio signal.

[0072] Fig. 9 According to the present invention Figure 8 A schematic diagram of the connection method of the microphone module in the embodiment.

[0073] exist Fig. 9 The arrangement positions of the microphone modules 110, 120 and 130 in the figure are in an isosceles right triangle. For example, the distance between the microphone 112 of the microphone module 110 and the microphone 122 of the microphone module 120 is d (e.g., 5 mm), and the distance between the microphone 112 of the microphone module 110 and the microphone 130 is also d, and the distance between the microphone 122 of the microphone module 120 and the microphone 130 is In this embodiment, the polar pattern with a heart shape represents the polar pattern of the differential audio signal generated by each microphone module, and its sensitivity scale unit is similar to Figure 5B In addition, a circle-like polar pattern (ie, having omnidirectionality) represents a polar pattern of the bypass audio signal generated by each microphone module.

[0074] For example, the directivity of the differential audio signal generated by the microphone module 110 is toward the left, the directivity of the differential audio signal generated by the microphone module 120 is toward the downward direction, and the directivity of the differential audio signal generated by the microphone module 130 is toward the upper right, which means that the directivity of each differential audio signal is perpendicular to the connection line between two adjacent microphone modules and toward the outside of the small array micro-electromechanical microphone device 100.

[0075] Fig.10 According to the present invention Figure 8 A schematic diagram of the connection method of the microphone module in the embodiment.

[0076] exist Fig.10 The microphone modules 110, 120 and 130 are arranged in an equilateral triangle, which means that the distance between the microphones of two adjacent microphone modules is d (e.g., 5 mm). In this embodiment, the polar pattern with a heart shape represents the polar pattern of the differential audio signal generated by each microphone module, and its sensitivity line unit is similar to Figure 5B In addition, the polar pattern similar to a circle (i.e., omnidirectional) represents the polar pattern of the bypass audio signal generated by each microphone module. For example, the directivity of the differential audio signal generated by microphone module 110 is toward the upper left, the directivity of the differential audio signal generated by microphone module 120 is toward the lower side, and the directivity of the differential audio signal generated by microphone module 130 is toward the upper right side, which means that the directivity of each differential audio signal is perpendicular to the line between two adjacent microphone modules and toward the outside of the small array micro-electro-mechanical microphone device 100.

[0077] Fig.11 FIG. 4 is a flow chart of a noise suppression method according to an embodiment of the present invention.

[0078] In step S1110 , a first microphone and a second microphone are used to extract a first audio signal and a second audio signal from a sound source respectively.

[0079] In step S1120, a first integrated circuit is used to perform a first logic operation on the first audio signal and the second audio signal to generate a first summed audio signal SUM, wherein the first summed audio signal SUM can be regarded as an omnidirectional audio signal, and has better sensitivity, noise floor and signal-to-noise ratio.

[0080] In step S1130, the first integrated circuit is used to sample and delay the second audio signal with the first clock signal, and the delayed second audio signal is subtracted from the first audio signal to obtain a first differential audio signal, and the first differential audio signal has a first directivity. The differential audio signal DMA can provide a beam function, for example, it can be regarded as an audio signal provided by a directional microphone in a microphone array of the first combination. In addition, the sampling delay of the differential audio signal DMA is sampled using a hardware circuit with a clock of 2.048MHz (unlimited). In addition to the microphone distance (e.g., 5mm) meeting the requirements of a small microphone array, the accuracy of the sampling delay and the fast / power-saving effect can also be achieved.

[0081] In step S1140, the second integrated circuit is used to bypass and output the second audio signal, and the second audio signal has omnidirectionality.

[0082] In summary, the present invention provides a small array micro-electromechanical microphone device and a noise suppression method, wherein the small array micro-electromechanical microphone device can provide a summed audio signal SUM, a differential audio signal DMA and a bypass audio signal BYPASS (e.g., MicB audio signal) to a back-end speech recognition device. The summed audio signal SUM can be regarded as an audio signal provided by an omnidirectional microphone in a first combination of microphone arrays, and compared with the sound signals XF and XB, the sensitivity of the summed audio signal SUM is increased by about 6dB (e.g., the signal correlation is 1), the background noise is increased by about 3dB (e.g., the signal correlation is 0), and the signal-to-noise ratio (SNR) is increased by about 3dB. The differential audio signal DMA can provide a beam function, for example, it can be regarded as an audio signal provided by a directional microphone in the first combination of microphone arrays. In addition, the sampling delay of the differential audio signal DMA is sampled using a hardware circuit with a clock of 2.048MHz (not limited). In addition to the microphone distance (e.g., 5mm) meeting the requirements of a small microphone array, the accuracy of the sampling delay and the fast / power saving effect can also be achieved. The BYPASS audio signal may also be regarded as an audio signal provided by the omnidirectional microphone in the microphone array of the first combination.

[0083] Although the present invention is disclosed as above in terms of preferred embodiments, it is not intended to limit the scope of the present invention. Those skilled in the art may make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined in the attached claims.

Claims

1. A small array micro-electromechanical microphone device, comprising: Mask; A circuit board coupled to the cover and comprising a first sound hole and a second sound hole; The first microphone module comprises: a first microphone and a first integrated circuit, wherein the first microphone extracts a first audio signal from a sound source via the first sound hole; and A second microphone module, comprising: a second microphone and a second integrated circuit, wherein the second microphone extracts a second audio signal from the sound source via the second sound hole, wherein the first microphone and the second microphone have the same sensitivity, phase and omnidirectionality, and the second audio signal is input to the first integrated circuit; The first integrated circuit performs a first logic operation on the first audio signal and the second audio signal to generate a first summed audio signal, wherein the first integrated circuit performs sampling delay on the second audio signal with a first clock signal, and subtracts the delayed second audio signal from the first audio signal to obtain a first differential audio signal, and the first differential audio signal has a first directivity. The second integrated circuit bypasses and outputs the second audio signal, and the second audio signal is omnidirectional. The first audio signal is further input to the second integrated circuit, and the second integrated circuit performs sampling delay on the first audio signal with the first clock signal, and subtracts the delayed first audio signal from the second audio signal to obtain a second differential audio signal, and the second differential audio signal has a second directivity different from the first directivity. The second integrated circuit performs a second logic operation on the second audio signal and the first audio signal to generate a second summed audio signal. The second integrated circuit further applies a first low-pass filter and a first equalizer, a first band-pass filter and a second equalizer, and a first high-pass filter and a third equalizer to the second summed audio signal, the second differential audio signal, and the first audio signal to generate a second low-frequency audio signal, a second intermediate-frequency audio signal, and a second high-frequency audio signal. The second integrated circuit further adds up the second low-frequency audio signal, the second intermediate-frequency audio signal and the second high-frequency audio signal to generate a second beam, and the second beam has the second directivity. 2 . The small array micro-electromechanical microphone device as claimed in claim 1 , wherein the first integrated circuit and the second integrated circuit are application-oriented integrated circuits, and the frequency of the first clock signal is 2.048 MHz. 3 . The small array micro-electromechanical microphone device as claimed in claim 2 , wherein a first distance between the first microphone and the second microphone is 5 mm.

4. The small array micro-electromechanical microphone device as described in claim 3, wherein the first integrated circuit further applies a second low-pass filter and a fourth equalizer, a second band-pass filter and a fifth equalizer, and a second high-pass filter and a sixth equalizer to the first summed audio signal, the first differential audio signal and the second audio signal respectively to generate a first low-frequency audio signal, a first intermediate-frequency audio signal and a first high-frequency audio signal, and the sensitivities of the first low-frequency audio signal, the first intermediate-frequency audio signal and the first high-frequency audio signal are substantially the same. 5 . The small array MEMS microphone device as claimed in claim 4 , wherein the passband of the low pass filter is below 200 Hz, the passband of the band pass filter is 200-7400 Hz, and the passband of the high pass filter is above 7400 Hz. 6 . The small array micro-electromechanical microphone device as claimed in claim 4 , wherein the first integrated circuit further sums the first low-frequency audio signal, the first intermediate-frequency audio signal and the first high-frequency audio signal to generate a first beam, and the first beam has the first directivity.

7. A noise suppression method for a small array micro-electro-mechanical microphone device, the small array micro-electro-mechanical microphone device comprising a mask; Circuit boards; A first microphone module includes: a first microphone and a first integrated circuit; and a second microphone module includes: a second microphone and a second integrated circuit, wherein the first microphone and the second microphone have the same sensitivity, phase and omnidirectionality, and the method includes: Utilizing the first microphone and the second microphone to extract a first audio signal and a second audio signal from a sound source respectively; Using the first integrated circuit to perform a first logic operation on the first audio signal and the second audio signal to generate a first summed audio signal; Using the first integrated circuit to perform sampling delay on the second audio signal with a first clock signal, and subtracting the delayed second audio signal from the first audio signal to obtain a first differential audio signal, wherein the first differential audio signal has a first directivity; and The second integrated circuit is used to bypass and output the second audio signal, wherein the second audio signal has omnidirectionality, The method further comprises: Using the second integrated circuit to perform a second logic operation on the second audio signal and the first audio signal to generate a second summed audio signal; Utilizing the second integrated circuit to sample and delay the first audio signal with the first clock signal, and subtracting the delayed first audio signal from the second audio signal to obtain a second differential audio signal, wherein the second differential audio signal has a second directivity different from the first directivity; utilizing the second integrated circuit to apply a first low-pass filter and a first equalizer, a first band-pass filter and a second equalizer, and a first high-pass filter and a third equalizer to the second summed audio signal, the second differential audio signal, and the first audio signal, respectively, to generate a second low-frequency audio signal, a second intermediate-frequency audio signal, and a second high-frequency audio signal; and The second integrated circuit is used to sum the second low-frequency audio signal, the second intermediate-frequency audio signal and the second high-frequency audio signal to generate a second beam, and the second beam has the second directivity. 8 . The noise suppression method as claimed in claim 7 , wherein the first integrated circuit and the second integrated circuit are application-oriented integrated circuits, and a frequency of the first clock signal is 2.048 MHz. 9 . The noise suppression method as claimed in claim 8 , wherein a first distance between the first microphone and the second microphone is 5 mm.

10. The noise suppression method according to claim 9, further comprising: The first integrated circuit is used to apply a second low-pass filter and a fourth equalizer, a second band-pass filter and a fifth equalizer, and a second high-pass filter and a sixth equalizer to the first summed audio signal, the first differential audio signal, and the second audio signal to generate a first low-frequency audio signal, a first intermediate-frequency audio signal, and a first high-frequency audio signal, respectively. The sensitivities of the first low-frequency audio signal, the first intermediate-frequency audio signal and the first high-frequency audio signal are substantially the same. 11 . The noise suppression method as claimed in claim 10 , wherein the pass band of the low pass filter is below 200 Hz, the pass band of the band pass filter is 200 to 7400 Hz, and the pass band of the high pass filter is above 7400 Hz.

12. The noise suppression method according to claim 10, further comprising: The first integrated circuit is further used to sum the first low-frequency audio signal, the first intermediate-frequency audio signal and the first high-frequency audio signal to generate a first beam, and the first beam has the first directivity.

13. A small array micro-electromechanical microphone device, comprising: Mask; A circuit board is coupled to the cover and includes a first sound hole, a second sound hole and a third sound hole; The first microphone module comprises: a first microphone and a first integrated circuit, wherein the first microphone extracts a first audio signal from a sound source via the first sound hole; and The second microphone module includes: a second microphone and a second integrated circuit, wherein the second microphone extracts a second audio signal from the sound source through the second sound hole, The third microphone module comprises: a third microphone and a third integrated circuit, wherein the third microphone extracts a third audio signal from the sound source via the third sound hole, wherein the first microphone, the second microphone and the third microphone have the same sensitivity, phase and omnidirectionality, Wherein, the first audio signal is input to the second integrated circuit, the second audio signal is input to the third integrated circuit, and the third audio signal is input to the first integrated circuit; The first integrated circuit is used to perform a first logic operation on the first audio signal and the third audio signal to generate a first summed audio signal, and to perform sampling delay on the third audio signal with a first clock signal, and to subtract the delayed third audio signal from the first audio signal to obtain a first differential audio signal, and the first differential audio signal has a first directivity. The first integrated circuit bypasses and outputs the first audio signal, and the first audio signal is omnidirectional. The second integrated circuit is used to perform a second logic operation on the second audio signal and the first audio signal to generate a second summed audio signal, and to perform sampling delay on the first audio signal using the first clock signal, and to subtract the delayed first audio signal from the second audio signal to obtain a second differential audio signal, and the second differential audio signal has a second directivity. The second integrated circuit bypasses and outputs the second audio signal, and the second audio signal is omnidirectional. The third integrated circuit is used to perform a third logic operation on the third audio signal and the second audio signal to generate a third summed audio signal, and to perform sampling delay on the second audio signal using the first clock signal, and to subtract the delayed second audio signal from the third audio signal to obtain a third differential audio signal, and the third differential audio signal has a third directivity. The third integrated circuit bypasses and outputs the third audio signal, and the third audio signal is omnidirectional. The first directivity is perpendicular to the line connecting the first microphone and the third microphone and is toward the outside of the small array micro-electro-mechanical microphone device, the second directivity is perpendicular to the line connecting the first microphone and the second microphone and is toward the outside of the small array micro-electro-mechanical microphone device, and the third directivity is perpendicular to the line connecting the second microphone and the third microphone and is toward the outside of the small array micro-electro-mechanical microphone device.

14. A small array micro-electromechanical microphone device as described in claim 13, wherein the settings of the first microphone, the second microphone and the third microphone form an isosceles right triangle, and the first distance between the first microphone and the second microphone is the first distance, and the first distance between the first microphone and the third microphone is the first distance*√2, and the first distance is 5 mm.

15. The small array MEMS microphone device as claimed in claim 13, wherein the first microphone, the second microphone and the third microphone are arranged in an equilateral triangle, and any two of the first microphone, the second microphone and the third microphone are at a first distance from each other, wherein the first distance is 5 mm.

Images