CN101790119A - Microphone array with second order directional pattern - Google Patents

Abstract

Disclose a kind of definite a plurality of microphones and whether had the frequency response characteristic of abundant coupling to be used for repeatedly pointing to the method for type microphone system.Also disclose the arrangement of microphone in this array of method determine a to(for) microphone array, described microphone array has at least three microphones, and one of them microphone is between other microphones.

Description

Microphone array with second order directional pattern

The application be that April 28, application number in 2004 are 200480011549.5 the applying date, denomination of invention divides an application for the Chinese invention patent application of " microphone array with second order directional pattern ".

Technical field

The present invention relates to have the microphone array of second order directional pattern.

Background technology

Can use two or more a plurality of isolated full sensing type microphone to constitute microphone array with order directional pattern.The system that uses two microphones to constitute an order directional pattern is widely used in the hearing aids at present.By use three or more microphones with form secondary or other more the order directional pattern of high order can improve pointing capability in theory.Yet, because the practical problem that must very closely mate of sensitivity of microphone, make these secondaries or more the pointing system of high order be difficult to the pointing capability that obtains to improve.Need to mate sensitivity microphones and also in the surplus method of improving directive property when sensitivity error is arranged still as well as possiblely.

Carried out such trial: measure the phase difference of microphone at the frequency place that just is lower than the microphone resonance frequency, and only accept the microphone group that the array of this phase difference in predetermined tolerance limit used.This trial is too limited to the microphone of finding out within the standard of falling into, and the microphone of abundant coupling is not still determined in this trial simultaneously.

Suppose that microphone not exclusively mates, also need to determine by what order microphone to be positioned in the array for the directive property of optimum.

The present invention has been proposed to solve these and other problems.

Summary of the invention

The purpose of an aspect of of the present present invention provides a kind of sensing type microphone system.

According to this aspect of the present invention, this system comprises the device that is used to provide the single order signal of representing a figure and is used for signal is carried out the device of low-pass filtering.This system also comprises device that is used to provide the second order signal of representing the secondary figure and the device that is used for the second order signal is carried out high-pass filtering.This system also comprises and being used for through the single order signal of low-pass filtering with through the device of the second order signal plus of high-pass filtering.

The purpose of another aspect of the present invention provides a kind of definite a plurality of microphones and whether has the frequency response characteristic of abundant coupling to be used for repeatedly pointing to the method for type microphone array.

According to this aspect of the present invention, determine the quality of microphone by resonance frequency and the Q that determines each microphone, and whether the difference between the resonance frequency of the Q of definite each microphone and each microphone falls within the acceptable tolerance limit in resonance peak zone coupling.

For example, usually, the frequency response of microphone has linear substantially part on a frequency range, at resonance frequency f _rBeing raised to peak value, is sloping portion afterwards.Difference between the size at the size of linear part and resonance frequency f place is commonly referred to Δ p.The Q of resonance is by being associated with Δ p: Δ p=20log Q, therefore mate Δ p and be equivalent to coupling Q.

Thus, determine the resonance frequency of Δ p He each microphone of each microphone.Whether the difference between the resonance frequency of the difference of the Δ p of definite each microphone and each microphone falls within the acceptable tolerance limit then.

For the microphone array with at least three microphones (one of them microphone is arranged between other microphones), another purpose of the present invention provides the layout of a kind of definite microphone in array to obtain the method for optimum directive property.

According to this aspect of the present invention, this method comprises by the order of the worst error in the sensing response of minimized array places microphone.Should place microphone in the following order: make responsiveness at central microphone on the major part of high band between the responsiveness of outmost microphone.In some cases, thus can determine this order by in the responsiveness of single frequency microphone being classified by microphone.

For example, measure the responsiveness of each microphone, and select to have the microphone of intermediate response degree as the microphone between other two microphones in the array at the frequency place that is higher than each microphone resonance frequency.

Description of drawings

Fig. 1 illustrates super heart (hypercardioid) figure and the secondary figure with high directivity;

Fig. 2 illustrates two pressure microphones;

Fig. 3 illustrates three pressure microphones;

Fig. 4 illustrates three order directional patterns;

Fig. 5 illustrates three second order directional patterns;

Fig. 6 is the block diagram that forms the circuit of bipolar (dipole) figure;

Fig. 7 is the block diagram that forms the circuit of super heart-shaped figure;

Fig. 8 is the block diagram that forms the circuit of four utmost points (quadrupole) figure;

Fig. 9 is the block diagram that forms the circuit of optimum secondary figure;

Figure 10 illustrates the curve chart of the sensitivity of full sensing type microphone, bipolar and four utmost points to frequency;

Figure 11 is the curve chart that the directional gain of a figure that is subjected to the little error effect of sensitivity of microphone is shown;

Figure 12 is the curve chart that the directional gain of the secondary figure that is subjected to the little error effect of sensitivity of microphone is shown;

Figure 13 illustrates the figure being subjected to the little error effect of sensitivity of microphone and the curve chart of secondary figure;

Figure 14 is the block diagram that mixes time sensing type system;

Figure 15 is the stereogram that is set to form two microphones of secondary figure;

Figure 16 is the block diagram to the realization of optimum secondary figure;

Figure 17 is the block diagram that provides the microphone array of second order directional pattern according to of the present invention;

Figure 18 is the frequency response curve of common microphone;

Figure 19 is the frequency response curve with three microphones of different high frequency response characteristics; And

Figure 20 is the frequency response curve with three microphones of different intermediate frequency response characteristics.

Embodiment

The present invention allows multiple multi-form embodiment, illustrate in the drawings and will describe the preferred embodiments of the present invention in detail at this, should understand the disclosure and should be considered as example, rather than be intended to wide aspect of the present invention is limited in the embodiment that illustrates principle of the present invention.

For easy to understand, be the nomenclature of some term used herein below:

Pressure microphone---normally used microphone type in hearing aids.This microphone detects the acoustic pressure of single-point.The pressure microphone has equal sensitivity to the sound from all directions.

The first difference figure---as the figure of the formation of the pressure differential between 2 in the space.The two-port microphone that often uses in the hearing aids is this type.

The second order difference figure---as the figure of the difference formation between two figures.

Bipolar---in forward direction and back to the response of side zero first difference figure to having identical response amplitude.On the mathematics, this figure has shape R (θ)=Bcos θ.

Cardioid---forward direction have the peak response degree and to the back to response only be zero first difference figure.Its graph function is R (θ)=A (1+cos θ).

Super cardioid---have the first difference figure of maximum sensitivity index.Its graph function is R (θ)=A (1+3cos θ).

Two-way---be used on fore-and-aft direction, having the adopted name of any figure of equal peak response degree.Bipolar is once two-way figure.Four utmost points are the two-way figures of secondary.

Four utmost points---graph function is R (θ)=Acos ²The two-way figure of the secondary of θ.

In hearing aids, provide significant benefits for the user aspect the hearing of adding directional microphone response pattern in noisy environment.At present, thus hearing aids manufacturer by making up two conventional microphones output otherwise provide order directional pattern by the figure that the figure with once sensing type microphone expands a conventional microphone.Under any situation, can obtain the scope (cardioid, super cardioid, two-way etc.) of an order directional pattern.In non-sensing noise range, these figures can provide the signal to noise ratio (snr) of maximum 6dB to increase.

By increasing the further improvement that other complexity of another grade can realize SNR in theory to sensing type system.The output of three conventional microphones of combination, the output of perhaps making up a pressure microphone and one or more a plurality of subgradient microphones can make SNR be increased to 9.5dB in theory.Provide below the theory of the performance that can obtain from the system with two and three pressure microphones is relatively estimated.Comprise to the system of the pressure microphone of or more a plurality of once sensing type microphone combination having similar performance, also will discuss it.Fig. 1 a shows super cardioid figure, and it is a highest figure of directive property.It is the highest and have a secondary figure of narrower response at forward direction that Fig. 1 b shows directive property.

Can be from the figure of two microphones acquisition

Provide two microphones of partition distance d as implied above, response R (θ) is provided by following formula:

$R\left(\mathrm{\&theta;}\right)=s_{-1}{e}^{-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}}+s_1{\mathrm{<figure-callout\; id="2"\; label="e\; j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">e</figure-callout>}}^{j\frac{\mathrm{kd}}{}}$

Wherein:

s _-1And s ₁Be the sensitivity of two microphones;

K=2 π/λ=2 π f/c are sound wave numbers;

λ is a wave length of sound;

F is a sound frequency;

C is the aerial speed of sound; And

θ is the angle between the line of connection microphone and the direction of propagation that enters wavefront (wavefront).

In hearing aids, microphone is at interval always much smaller than wavelength, so kd＜＜1.For the responsiveness of an order directional pattern is estimated, only need to remain into the once item of kd.So, response formula can be expanded into:

$R\left(\mathrm{\&theta;}\right)\&ap;s_{-1}(1-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;})$

$+s_1(1+j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;})$

$\&ap;(s_{-1}+s_1)+j\frac{\mathrm{kd}}{2}(s_1-s_{-1})\cos \mathrm{\&theta;}$

$\&ap;A+B\cos \mathrm{\&theta;}$

The graphical-set that A and B can obtain for real number numerical value is limacon (limacon) graphical-set.The example of this family has been shown among Fig. 4.Notice that " forward direction " direction is to the right a direction among the figure.

Fig. 2 illustrates two microphones, they can provide bipolar (Fig. 4 a), the first difference order directional pattern of cardioid (Fig. 4 b) and super cardioid (Fig. 4 c).

When A=0, form bipolar figure.Bipolar on to the direction of side its response be zero.Second graph is a cardioid, forms when A=B.This cardioid in the back to only being zero.The 3rd figure is super cardioid, forms when B=3A.Super cardioid is a highest figure of directional gain.

Can be from the figure of three microphones acquisition

Provide three microphones of partition distance d as shown in Figure 3, response R (θ) is provided by following formula:

$R\left(\mathrm{\&theta;}\right)=s_{-1}{e}^{-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}}+{\mathrm{<figure-callout\; id="0"\; label="s"\; filenames="HSA00000029109400011.png,HSA00000029109400022.png"\; state="\{\{state\}\}">s</figure-callout>}}_0+s_1{e}^{j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;}}$

Wherein:

s _-1, s ₀And s ₁Be the sensitivity of microphone;

K=k=2 π/λ=2 π f/c are sound wave numbers;

λ is a wave length of sound;

F is a sound frequency;

C is the aerial speed of sound; And

θ is the angle between the line of connection microphone and the direction of propagation that enters wavefront.

As mentioned above, in hearing aids, microphone is at interval always much smaller than wavelength, thus kd＜＜1.Estimate for response, must remain into the quadratic term of kd second order directional pattern.So, response formula can be expanded into:

$R\left(\mathrm{\&theta;}\right)\&ap;s_{-1}(1-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}-\frac{{\left(\mathrm{kd}\right)}^2}{8}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})+{\mathrm{<figure-callout\; id="0"\; label="s"\; filenames="HSA00000029109400011.png,HSA00000029109400022.png"\; state="\{\{state\}\}">s</figure-callout>}}_0$

$+s_1(1+j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;}-\frac{{\left(\mathrm{kd}\right)}^2}{8}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})$

$\&ap;({\mathrm{<figure-callout\; id="0"\; label="s"\; filenames="HSA00000029109400011.png,HSA00000029109400022.png"\; state="\{\{state\}\}">s</figure-callout>}}_0+s_{-1}+s_1)+j\frac{\mathrm{kd}}{2}(s_{-1}+s_1)\cos \mathrm{\&theta;}$

$-\frac{{\left(\mathrm{kd}\right)}^2}{8}(s_{-1}+s_1){\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;}$

$\&ap;A+B\cos \mathrm{\&theta;}+\mathrm{<figure-callout\; id="2"\; label="C\; cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">C</figure-callout>}{\cos }^{}{}^2\mathrm{\&theta;}$

The example of this family has been shown among Fig. 5, and it shows four utmost point figures, and (Fig. 5 a) and other two figures.Notice that " forward direction " direction is to the right a direction among the figure.

When A=B=0, form four utmost point figures.Four utmost points are zero in its response of direction to the side.When A=0 and B=C, form second graph.This figure is set to the back to being zero.When B=2A and C=5A, form the 3rd figure.This is the highest secondary figure of directional gain.

Directional gain

Check the above order directional pattern of two microphone systems and three Mike's wind systems, obviously, some figures have the response pattern of broad at forward direction, and other figures have more inhibition in the back to direction.A kind of method that the directive property of different graphic is compared is the tolerance that is called directional gain (DI).DI is the ratio at sound signal that is received by full sensing from the sound field that all directions arrive in the same manner and the signal that is received by order directional pattern, is unit with dB.On the mathematics, directional gain DI is provided by following formula:

$\mathrm{DI}=10\log \left\{\frac{{2\left[R\left(0\right)\right]}^2}{\underset{0}{\overset{\mathrm{\&pi;}}{\&Integral;}}{\left[R\left(\mathrm{\&theta;}\right)\right]}^2\sin \mathrm{\&theta;d\&theta;}}\right\}$

Notice that this is one and is easy to idealized tolerance that Utopian microphone figure is calculated.Be arranged in hearing aids and be installed in the actual conditions of user's head at microphone, the figure height is asymmetric and be difficult to calculate DI.In addition, Utopian even sound field seldom is true to nature approximate for the actual environment noise that has in the true environment.Yet DI remains the useful metrics that system is compared.

The DI of two microphones

To a beam pattern (beam pattern), replace above-mentioned formula:

$\mathrm{DI}=10\log \left[\frac{2{(A+B)}^2}{\underset{0}{\overset{\mathrm{\&pi;}}{\&Integral;}}{(A+B\cos \mathrm{\&theta;})}^2\sin \mathrm{\&theta;d\&theta;}}\right]$

$=10\log \left[\frac{{(A+B)}^2}{{\mathrm{<figure-callout\; id="2"\; label="A"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">A</figure-callout>}}^2+\frac{1}{3}{B}^2}\right]$

Following table has been listed the DI of the several figures in the limacon family.The figure that is called super cardioid is optimum on such meaning: its have the highest directive property in the figure once.

Figure	??A	??B	??DI
				The full sensing	??1.0	??0.0	??0.0
Bipolar	??0.0	??1.0	??4.8
				Cardioid	??0.5	??0.5	??4.8
Super cardioid	??.25	??.75	??6.0

The concept nature of two microphones realizes

In order to reach actual realization, need calculate the addition coefficient of microphone from the value of the A of desirable figure and B.From above formula, A and B are defined as:

A＝s _-1+s ₁

$B=j\frac{\mathrm{kd}}{2}(s_1-s_{-1})$

Addition coefficient to microphone is found the solution:

${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">s</figure-callout>}}_1=\frac{1}{2}A-\frac{j}{\mathrm{kd}}B$

$s_{-1}=\frac{1}{2}A+\frac{j}{\mathrm{kd}}B$

As example, can consider to form the block diagram of bipolar figure.For bipolar:

A＝0，B＝1

${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">s</figure-callout>}}_1=-\frac{j}{\mathrm{kd}},$ $s_{-1}=\frac{j}{\mathrm{kd}}$

Realize shown in Fig. 6 that this points to the block diagram of handling.The integration filter of output place must be to providing the flat frequency response from bipolar signal.This carries out signal plus to finish the work with single filter before being implemented in filtering.

More complete example is to form an optimum figure, promptly super cardioid.For this figure:

$A=\frac{1}{4},$ $B=\frac{3}{4}$

${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">s</figure-callout>}}_1=\frac{1}{8}-j\frac{3}{4\mathrm{kd}},$ $s_{-1}=\frac{1}{8}+j\frac{3}{4\mathrm{kd}}$

The block diagram of realizing that this sensing is handled has been shown among Fig. 7, and it is that the block diagram that forms the required circuit of super cardioid figure is shown.

The DI of three microphones

To the secondary beam pattern, replace above-mentioned formula:

$\mathrm{DI}=10\log \left[\frac{2{(A+B+C)}^2}{\underset{0}{\overset{\mathrm{\&pi;}}{\&Integral;}}{(A+B\cos \mathrm{\&theta;}+\mathrm{<figure-callout\; id="2"\; label="C\; cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">C</figure-callout>}{\cos }^{}{}^2\mathrm{\&theta;})}^2\sin \mathrm{\&theta;d\&theta;}}\right]$

$=10\log \left[\frac{{(A+B+C)}^2}{{\mathrm{<figure-callout\; id="2"\; label="A"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">A</figure-callout>}}^2+\frac{1}{3}{\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">B</figure-callout>}}^2+\frac{1}{5}{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">C</figure-callout>}}^2+\frac{2}{3}\mathrm{AC}}\right]$

Following table has been listed the DI of several secondary figures.The figure of listing as optimum secondary is optimum on such meaning: it has the highest directive property in all secondary figures.

Figure	??A	??B	??C	??DI
					The full sensing	??1.0	??0.0	??0.0	??0.0
Four utmost points	??0.0	??0.0	??1.0	??7.0
					Behind the W/ to zero	??0.5	??0.5	??0.5	??8.8

Figure	??A	??B	??C	??DI
					Optimum secondary	??-1/6	??1/3	??5/6	??9.5

The concept nature of three microphones realizes

In order to reach actual realization, need calculate the addition coefficient of microphone from the value of A, the B of desirable figure and C.According to above formula, A, B and C are defined as:

A＝s ₀+s _-1+s ₁

$B=j\frac{\mathrm{kd}}{2}(s_1-s_{-1})$

$\frac{{\left(\mathrm{kd}\right)}^2}{8}({\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">s</figure-callout>}}_1+s_{-1})$

The microphone addition coefficient is found the solution:

${\mathrm{<figure-callout\; id="0"\; label="s"\; filenames="HSA00000029109400011.png,HSA00000029109400022.png"\; state="\{\{state\}\}">s</figure-callout>}}_0=A+\frac{8}{{\left(\mathrm{kd}\right)}^2}C$

${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">s</figure-callout>}}_1=-\frac{4}{{\left(\mathrm{kd}\right)}^2}C-\frac{j}{\mathrm{kd}}B$

$s_{-1}=-\frac{4}{{\left(\mathrm{kd}\right)}^2}C+\frac{j}{\mathrm{kd}}B$

As example, consider the block diagram of Fig. 8, it can form four utmost point figures.For four utmost points:

A＝0，B＝0，C＝1

${\mathrm{<figure-callout\; id="0"\; label="s"\; filenames="HSA00000029109400011.png,HSA00000029109400022.png"\; state="\{\{state\}\}">s</figure-callout>}}_0=\frac{8}{{\left(\mathrm{kd}\right)}^2},$ ${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">s</figure-callout>}}_1=-\frac{4}{{\left(\mathrm{kd}\right)}^2},$ $s_{-1}=-\frac{4}{{\left(\mathrm{kd}\right)}^2}$

The double integration filter of output place must provide the flat frequency response to the signal from four utmost points.This carries out signal plus to finish the work with single filter before being implemented in filtering.

More complete example is to form optimum secondary figure.For this figure:

$A=-\frac{1}{6},$ $B=\frac{1}{3},$ $C=\frac{5}{6},$ ${\mathrm{<figure-callout\; id="0"\; label="s"\; filenames="HSA00000029109400011.png,HSA00000029109400022.png"\; state="\{\{state\}\}">s</figure-callout>}}_0=-\frac{1}{6}+\frac{20}{3{\left(\mathrm{kd}\right)}^2},$

${\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">s</figure-callout>}}_1=-\frac{10}{3{\left(\mathrm{kd}\right)}^2}-\frac{j}{3\mathrm{kd}},$ $s_{-1}=-\frac{10}{3{\left(\mathrm{kd}\right)}^2}+\frac{j}{3\mathrm{kd}}$

The block diagram of realizing that this sensing is handled has been shown among Fig. 9, and it is that the block diagram that forms the required circuit of optimum secondary figure is shown.

Sensitivity of microphone error in figure

Relatively at Fig. 7 of a figure with at Fig. 9 of secondary figure, the complexity that seems to be used for the circuit of aftertreatment is not big especially.Yet surperficial simply may be fraudulent, because the tolerance limit of component value (comprising sensitivity of microphone) is much bigger.

More than analyze and supposed that the sensitivity of two microphones is identical, and with the addition coefficient in the unlimited precision realization treatment circuit.Never be such situation in practice.The sensitivity of microphone always has some variations aborning.Certainly in production technology, manual measurement and coupling are carried out in sensitivity, and in hearing aids, poor sensitivity is compensated in real time automatically.However, still always have some residual errors.The influence of sensitivity error to beam pattern shape and directional gain will be examined or check in this part.

Because this problem only relates to the poor of sensitivity, so hypothesis microphone s ₁Sensitivity be correct, | s _-1The incorrect degree of sensitivity be mark δ |.So this graphics calculations is:

$R\left(\mathrm{\&theta;}\right)=s_{-1}(1+\mathrm{\&delta;})e^{-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}}+s_1{e}^{j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;}}$

$\&ap;(\frac{A}{2}+\frac{\mathrm{jB}}{\mathrm{kd}})(1+\mathrm{\&delta;})(1-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;})+(\frac{A}{2}-\frac{\mathrm{jB}}{\mathrm{kd}})(1-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;})$

$\&ap;(A+B\cos \mathrm{\&theta;})+\frac{\mathrm{\&delta;}}{2}(A+B\cos \mathrm{\&theta;})+j\frac{\mathrm{\&delta;B}}{\mathrm{kd}}$

More than first be ideal response.Suppose δ＜＜1, then second very little.In addition, second has desirable directive property, so it can not make the directive property deterioration of figure.Yet the 3rd does not have desirable directive property, and may be not little.Suppose frequency kd＜＜1 in all concerns before.Yet,, influence more obvious at the low frequency place.Inevitably, there is a frequency, in response, accounts for superiority being lower than this most last error term more than frequency place.

Sensitivity of microphone error in the secondary figure

Analyze and supposed that also the sensitivity of three microphones is identical, and hypothesis is with the addition coefficient in the unlimited precision realization treatment circuit.This never is actual conditions equally.

Because this problem only relates to the poor of sensitivity, therefore suppose microphone s ₀Sensitivity be correct, s _-1And s ₁The incorrect degree of sensitivity be mark δ ₁And δ ₁So this graphics calculations is:

$R\left(\mathrm{\&theta;}\right)=s_{-1}{e}^{-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}}+{\mathrm{<figure-callout\; id="0"\; label="s"\; filenames="HSA00000029109400011.png,HSA00000029109400022.png"\; state="\{\{state\}\}">s</figure-callout>}}_0+s_1{e}^{j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;}}$

$\&ap;(1+{\mathrm{\&delta;}}_{-1})(-\frac{4}{{\left(\mathrm{kd}\right)}^2}C+\frac{j}{\mathrm{kd}}B)e^{-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}}+(A+\frac{8}{{\left(\mathrm{kd}\right)}^2}C)$

$+(1+{\mathrm{\&delta;}}_1)(-\frac{4}{{\left(\mathrm{kd}\right)}^2}C-\frac{j}{\mathrm{kd}}B)e^{j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;}}$

$\&ap;(1+{\mathrm{\&delta;}}_{-1})(-\frac{4}{{\left(\mathrm{kd}\right)}^2}C+\frac{j}{\mathrm{kd}}B)(1-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}-\frac{{\left(\mathrm{kd}\right)}^2}{8}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})$

$+(A+\frac{8}{{\left(\mathrm{kd}\right)}^2}C)$

$\&ap;(1+{\mathrm{\&delta;}}_{-1})(-\frac{4}{{\left(\mathrm{kd}\right)}^2}C-\frac{j}{\mathrm{kd}}B)(1+j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;}-\frac{{\left(\mathrm{kd}\right)}^2}{8}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})$

$\&ap;(A+B\cos \mathrm{\&theta;}+\mathrm{<figure-callout\; id="2"\; label="C\; cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">C</figure-callout>}{\cos }^{}{}^2\mathrm{\&theta;})+\frac{1}{2}({\mathrm{\&delta;}}_1+{\mathrm{\&delta;}}_{-1})(B\cos \mathrm{\&theta;}+\mathrm{<figure-callout\; id="2"\; label="C\; cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">C</figure-callout>}{\cos }^{}{}^2\mathrm{\&theta;})$

$+j\frac{({\mathrm{\&delta;}}_{-1}-{\mathrm{\&delta;}}_1)}{\mathrm{kd}}B-\frac{4({\mathrm{\&delta;}}_{-1}+{\mathrm{\&delta;}}_1)}{{\left(\mathrm{kd}\right)}^2}C$

More than first be desirable response.Suppose δ＜＜1, then second very little, so it can not make the directive property deterioration of figure.Yet remaining item does not have desirable directive property, and may be not little.The 3rd is the once item of kd, and is equivalent to the error in the figure one time.The most last error term is the quadratic term of kd, and at the low frequency place figure is had bigger influence.Be to begin with hypothesis in the kd of the frequency place of all concerns＜＜1.Yet,, influence more obvious at the low frequency place.Inevitably, have a frequency, in response, account for superiority being lower than this most last error term more than frequency place, and this frequency is higher than the frequency that causes problem for figure.

The sensitivity of order directional pattern and noise

Form once with the second order directional pattern process in, deduct signal from microphone, this has reduced the output-voltage levels of wave beam significantly.Figure 10 illustrates the output sensitivity of the sensing wave beam of comparing with the sensitivity that is used to form the full sensing type microphone that points to wave beam.In order to illustrate, the original microphone that illustrates has (Itasca, IL, US) the similar frequency response of the frequency response of EM microphone series with Knowles Electronics LLC.Yet any other microphone family should show similar performance.With respect to single microphone, the sensitivity of once bipolar figure (middle curve) drops on 6dB/ octave (octave), its output 20dB under single microphone at the 500Hz place.Other figure has approximately uniform sensitivity and descends.With respect to single microphone, secondary four utmost point figures (following curve) drop on the 12dB/ octave, the low 40dB at the 1kHz place.

The internal noise of wave beam is the noise power sum from the microphone that is used to form this wave beam.In bipolar figure, internal noise is than the high 3dB of the noise in the single microphone.In four utmost point figures, internal noise is than the high 4.8dB of single microphone.Consider noise itself, these noise increases are not big shortcomings.Yet they combine with the sensitivity reduction and produce potential problem.

Reason is in great majority are used, and will use the decline of bigger gain with compensating signal sensitivity at the low frequency place.The sensitivity of this gain recovery signal, but also the low frequency internal noise has been amplified the identical factor.For bipolar figure, this will make the following internal noise of 500Hz increase above 20dB, and for four utmost point figures, it will make the following noise of 1kHz increase more than the 40dB.

For a figure, it is acceptable that this noise increase has only been covered in the noisy environment of internal noise in the high level ambient noise.In quiet environment, hearing aids should switch to the pattern of the more quiet full sensing type microphone of use.For the secondary figure, EQ Gain increases so many noise below 1kHz so that to use this figure at lower frequency may be unpractical.

In addition, for the secondary figure, also has a problem that limits its performance below 1kHz.This problem is discussed below.

Here the example of Ti Chuing relates to three microphone arrays that total length is 10mm.Use instruction of the present invention also can design the array of other sizes.Array for long can expand to the use to the secondary figure frequency lower than described example.For short array, the crossover frequency between single treatment and the aftertreatment need appear at higher frequency.These influences are included in the design formula by the factor kd that comprises array length.

Frequency limitation than high order directive property

Above formula shows at very low frequency place, and the inevitable little deviation of sensitivity of microphone will make the serious deterioration of graphics shape.Important problem is: this deterioration becomes problem in what frequency.

First example has been shown among Figure 11, and it shows for a figure directional gain that is subjected to the little error effect of sensitivity of microphone and descends at the low frequency place.In first example, investigated figure of optimum that the microphone by a pair of 10mm of being separated by, approximate match forms, promptly super cardioid.In this example, the sensitivity error δ of permission 0.05.This roughly is that the amplitude of half dB does not match or 3.5 ° phase error.Super cardioid figure has the desirable directive property of 6dB.When comprising sensitivity error, the directive property limiting value that this ideal directive property is high frequency treatment.This illustrate DI at the low frequency place deterioration how.For this example, DI drops to 5dB at 500Hz, drops to 4dB at 250Hz.For than this little DI value, this curve chart may be inaccurate.That uses is approximate only effective to the sensitivity error of smaller value.Hope obtains high DI value on the wide region of correlated frequency.

Second example shown in Figure 12, its directive property index that shows for the secondary figure that is subjected to sluggishness error (5%) influence may be little of not accepting at whole audio bandwidth.In this second example, consider the Quadratic Optimum figure.For three microphones are assemblied in the space available in the hearing aids, total with three microphones | bore | (aperture) remain on 10mm.If allow sensitivity error to have the amplitude identical with the front, then DI changes along with frequency as illustrated in fig. 12.At this other sensitivity error place of level, the value of secondary figure is very little.Except the frequency more than the 2800Hz, the directive property index of secondary figure is no more than the directive property index of a figure, and is higher than 5Hz up to frequency, and DI just reaches its total head.

Make the secondary figure to be with necessary several respects:

● only the frequency that is higher than 1kHz is used the secondary figure.This makes that the phase matched of sensitivity of microphone is more approaching.

● use the microphone that 10kHz is at least had flat response.

● comprise automatic self adaptation amplitude matches circuit.

Preceding two features provide smooth microphone frequency response in the whole bandwidth of using the secondary figure.This means for two microphone phase responses very near zero, and eliminated the unmatched any degree of freedom of microphone phase place.Any of sensitivity amplitude that the 3rd feature has automatically compensated two microphones do not match or drifts about.

According to these hypothesis, can make do not match (being δ) of microphone to drop to 0.01.Figure 13 shows at figure of low frequency use and at high frequency and uses the secondary figure that the hybrid pointing figure of the DI with improvement is provided.Himself can not use the secondary figure.Below 1kHz, it is so big so that can not look to second order directional that pattern error becomes.Yet, by using secondary figure, the average DI that can obtain to increase than figure of low frequency use and in higher-frequency.Such a hybrid system can utilize the secondary figure high-frequency range than high directivity, provide acceptable directive property at lower frequency simultaneously.Figure 13 illustrates the DI of super cardioid figure and secondary figure.Hybrid system attempts to reach the higher DI of directive property in two figures in each frequency.

The concept nature of hybrid pointing type system realizes

Figure 14 is the block diagram of hybrid pointing type system.At first Wai Mian two microphones make its signal gain be adjusted to amplitude matches with middle microphone.Combined microphone signal is to form an optimum figure and optimum secondary figure simultaneously then.At last, figure is carried out filtering and it is made up so that output comprises from the high frequency of secondary figure with from the mode of the low frequency of a figure.

Also has the additional design feature that to improve second order directional.Can be so that the residue matching error after regulating for two microphones has the gain adjusting circuit that the mode of opposite symbol designs two microphones in outside.In other words, δ _-1Have and δ ₁Opposite symbol.If so do, the composition of pattern error maximum then, promptly May be littler.If this makes this value can reduce 2 the factor, then can improve the DI of hybrid pointing type system significantly.The curve chart of Figure 14 shows this situation.

Use the secondary of once sensing type microphone to realize

As using three pressure microphones to obtain the alternative arrangement of second order directional, can also use two once sensing type microphones.Figure 15 illustrates the arrangement of two this microphones, and the port spacing distance is that each microphone of d/2 is placed end-to-end so that the total linear spacing of end port is d.The advantage of this realization is because difference is to pass the beat of single diaphragm, does not therefore have sensitivity error in the figure of the directional microphone that separates.Therefore, this figure sensitivity error for once.

Begin if having the hypothesis of bipolar figure with each microphone, then total microphone response is:

$R\left(\mathrm{\&theta;}\right)=\mathrm{<figure-callout\; id="1"\; label="j\; kdB"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{{\mathrm{kdB}}_{}}{}\frac{{}_1}{2}\cos \mathrm{\&theta;}{e}^{-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}}+\mathrm{<figure-callout\; id="2"\; label="j\; kdB"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{{\mathrm{kdB}}_{}}{}\frac{{}_2}{2}\cos \mathrm{\&theta;}{e}^{j\frac{\mathrm{kd}}{1}\cos \mathrm{\&theta;}}$

$\&ap;\mathrm{<figure-callout\; id="1"\; label="j\; kdB"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{{\mathrm{kdB}}_{}}{}\frac{{}_1}{2}\cos \mathrm{\&theta;}(1-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;})+\frac{{\mathrm{<figure-callout\; id="2"\; label="kdB"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">kdB</figure-callout>}}_2}{2}\cos \mathrm{\&theta;}(1+j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;})$

$\&ap;j\frac{\mathrm{kd}}{2}({\mathrm{<figure-callout\; id="1"\; label="B"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">B</figure-callout>}}_1+B_2)\cos \mathrm{\&theta;}+\frac{{\left(\mathrm{kd}\right)}^2}{4}(B_1-B_2){\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;}$

Here, factor jkd/2 is included in the sensitivity of each microphone to clearly illustrate the frequency response of final graphics.If but two bipolar microphones have the direction of identical axial sensitivity directed in opposite, then:

B ₂=-B ₁=B, and

$R\left(\mathrm{\&theta;}\right)\&ap;\frac{{\left(\mathrm{kd}\right)}^2\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">B</figure-callout>}}{2}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;S}$

Perhaps

$R\left(\mathrm{\&theta;}\right)\&ap;\frac{{\left(\mathrm{kd}\right)}^2\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">B</figure-callout>}}{2}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;}+j\frac{\mathrm{\&delta;<figure-callout\; id="2"\; label="kdB"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">kdB</figure-callout>}}{2}\cos \mathrm{\&theta;}$

In if sensitivity error is included in.This its error aspect that is implemented in has two advantages compared with the version of front.At first, error term only has a factor kd less than figure.Therefore the second, this error term has bipolar shape, destroys on to the direction of side less.Note as yet not to carrying out any explanation with any departing from of perfect condition in two bipolar graphics shapes.It may add the additive error of the obvious advantage that is enough to offset this realization potentially.

The another kind of possibility of sensing type microphone is to use its internal latency parameter to be adjusted to the first difference microphone that provides cardioid figure shape.So have:

$R\left(\mathrm{\&theta;}\right)=\mathrm{<figure-callout\; id="1"\; label="j\; kdB"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{{\mathrm{kdB}}_{}}{}\frac{{}_1}{2}(1+\cos \mathrm{\&theta;})e^{-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;}}+\mathrm{<figure-callout\; id="2"\; label="j\; kdB"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{{\mathrm{kdB}}_{}}{}\frac{{}_2}{2}(1+\cos \mathrm{\&theta;})e^{j\frac{\mathrm{kd}}{1}\cos \mathrm{\&theta;}}$

$\&ap;\mathrm{<figure-callout\; id="1"\; label="j\; kdB"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{{\mathrm{kdB}}_{}}{}\frac{{}_1}{2}(1+\cos \mathrm{\&theta;})(1-\mathrm{<figure-callout\; id="2"\; label="j\; kd"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{\mathrm{kd}}{}\frac{}{2}\cos \mathrm{\&theta;})+\frac{{\mathrm{<figure-callout\; id="2"\; label="kdB"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">kdB</figure-callout>}}_2}{2}(1+\cos \mathrm{\&theta;})(1+j\frac{\mathrm{kd}}{2}\cos \mathrm{\&theta;})$

$\&ap;j\frac{\mathrm{kd}}{2}({\mathrm{<figure-callout\; id="1"\; label="B"\; filenames="HSA00000029109400012.png,HSA00000029109400013.png"\; state="\{\{state\}\}">B</figure-callout>}}_1+B_2)(1+\cos \mathrm{\&theta;})+\frac{{\left(\mathrm{kd}\right)}^2}{4}(B_1-B_2)(\cos \mathrm{\&theta;}+{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})$

If also allow B ₂=-B ₁=B, then:

$R\left(\mathrm{\&theta;}\right)\&ap;\frac{{\left(\mathrm{kd}\right)}^{2}B}{2}(\cos \mathrm{\&theta;}+{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})+j\frac{\mathrm{\&delta;kdB}}{2}(1+\cos \mathrm{\&theta;})$

This is the secondary figure that paint the front, its in the back to being zero and having the desirable DI of 8.8dB.

The figure that is formed by two sensing type microphones with maximum possible directive property has angular response

$R\left(\mathrm{\&theta;}\right)\&ap;\frac{{\left(\mathrm{kd}\right)}^2}{2}(\frac{3}{8}\cos \mathrm{\&theta;}+\frac{5}{8}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})$

This figure has the desirable DI of 9.0dB.It is to be formed by two following figures of angular response:

$R\left(\mathrm{\&theta;}\right)\&ap;\frac{\mathrm{kd}}{2}(\frac{3}{8}+\frac{5}{8}\cos \mathrm{\&theta;})$

Secondary figure with optimum directive property also can be added full sensing type microphone by two sensing type microphones and be formed.

Last example shown in Figure 16 is the realization block diagram of optimum secondary figure.Consider to form optimum secondary figure.The front illustrates it and has graph function

In this case, utilization also has full sensing type microphone so that the minimum fact of the noise characteristic in the quiet environment except two once sensing type microphones.This microphone that is placed on the acoustic centre of source can very directly provide the major event in the graph function.Right latter two points to item can come from two identical microphones.If each sensing type microphone all has figure

And deduct the output signal of two microphones, then these two figure alone is:

$R\left(\mathrm{\&theta;}\right)=\frac{{\left(\mathrm{kd}\right)}^2}{2}(\frac{2}{7}\cos \mathrm{\&theta;}+\frac{5}{7}{\mathrm{<figure-callout\; id="2"\; label="cos"\; filenames="HSA00000029109400011.png,HSA00000029109400032.png"\; state="\{\{state\}\}">cos</figure-callout>}}^2\mathrm{\&theta;})$

It is added the pressure microphone to form final graphics.

Sensing type microphone array 10 with first, second and the 3rd full sensing type microphone 12,14 and 16 has been shown among Figure 17.The typical frequencies response curve of microphone has been shown among Figure 18.Usually, this frequency response has overall linear part 18, at resonance frequency f _rRising to peak value 20, is sloping portion 22 afterwards.As mentioned above, preferably, all microphones in the array all have identical response characteristic on the gamut of correlated frequency.But usually, this is not a viable commercial in practice.Therefore, have been found that the key property that will pay close attention to is decay and the microphone with similar attenuation characteristic mated.

A kind of method that microphone with similar attenuation characteristic is mated is that (Δ p is the amplitude and the resonance frequency f of linear part 18 by measuring (1) its Δ p _rPoorly between the amplitude at place) and the resonance frequency f of (2) each microphone _rWhether fully the directive property index of finally can accepting according to hope is identified for determining two microphones tolerance limit of coupling.As long as each Δ p and the resonance frequency f of three microphones _rBetween difference within predetermined tolerance limit, be acceptable with regard to will be understood that three microphones 12,14 and 16 for specific array.

Also can use other standards to determine whether microphone has the attenuation characteristic of enough couplings.

For example, can use from resonance frequency f _rThe point of following 3dB and resonance frequency f _rPoint between the measurement of difference on the frequency (being called Δ f).Alternatively, can use divided by resonance frequency f _rΔ f (being also referred to as the Q of resonance).The all or else same angle of in these methods each has provided information similar.

The Q of resonance is approximate to be associated with Δ p, and wherein Δ p approximates 20logQ, and therefore coupling Δ p is equivalent to coupling Q between microphone.

In case determined that three specific microphones are acceptable for specific array, then also will select microphone to be placed in the array with what order.Observe the formula of sensitivity of microphone error in the above-mentioned secondary figure, see that last is the worst error item, because the product kd in the denominator is very little, and with square increase of frequency.Microphone 12,14 and 16 can be arranged in the array so that the amplitude of worst error item minimizes on the working band of array.Mark δ ₁Be the error of an outside microphone, mark δ _-1It is the error of another outside microphone.If δ ₁And δ _-1Opposite in sign, they are partly cancelled each other.Though on practical significance, can not make error just in time equate and opposite, opposite by making them at least, reduced the amplitude of whole error term.Might mark δ ₁And δ _-1Be not all opposite in all frequencies, that is, the response amplitude curve may intersect.Because error term increases rapidly with frequency, cancel each other at the highest frequency place of expectation array work so the most important thing is mark.Usually, closely the microphone of coupling has in the resonance peak zone response amplitude that intersects at the most once, and not so intersection point keeps almost parallel near resonance frequency.This is hinting in resonance frequency and fully is being lower than or fully is higher than under the situation of maximum operating frequency of array, can use simple method to find out optimum microphone order.

Fully be lower than the situation of the maximum operating frequency of array for the resonance frequency of microphone, sloping portion of the response curve of these three microphones by observing array 10 is realized.With reference to Figure 19, usually, sloping portion 22a, 22b and the 22c of three microphones are substantially parallel.So at test frequency f _r(this frequency is higher than the resonance frequency of each microphone) observes the relative amplitude of each curve.Selection has the microphone of intermediate response amplitude as middle microphone 14, and in addition two be the microphone 12 and 16 of outside.

In addition, find that aforesaid coupling to Q and Δ p has at their resonance peak zone almost parallel or at the response curve that intersects near the zone of peak maximum near zone.

Therefore, as shown in Figure 20, the microphone that has intermediate response at the sloping portion of response curve also has intermediate response at the rising part of curve.Therefore, also can use the frequency of rising part of response as the equivalent standard of microphone in the middle of selecting.

Though specific embodiment is illustrated and illustrates, under the situation that does not significantly break away from spirit of the present invention, can expect a lot of modifications, thereby protection range is only limited by the scope of claims.

Claims (7)

1. determine the method for the arrangement of microphone in this array for a microphone array for one kind, described microphone array has at least three microphones, and one of them microphone is set between other microphones, and the method includes the steps of:

Measure of the response of each microphone at the frequency place of the resonance frequency that is lower than each microphone; And

Selection has the microphone of intermediate response as the microphone between other microphones in the array.

2. method that is used for selecting to be used for the microphone of microphone array from a plurality of microphones, each microphone in wherein said a plurality of microphones all has the response curve that is associated with it on a frequency range, and the method includes the steps of:

Determine each the response characteristic in described a plurality of microphone, this response characteristic is the measurement to the response at the frequency place near the response region the microphone substantially parallel resonance response of corresponding response curve therein; And so that the mode of the response characteristic of each in first microphone and second microphone within predictive error value each other selects first microphone and second microphone from described a plurality of microphones.

3. according to the method for claim 2, also comprise the step of from described a plurality of microphones, selecting the 3rd microphone, the 3rd microphone has the response characteristic between each the response characteristic in first microphone and second microphone.

4. according to the method for claim 3, also comprise so that the mode of the 3rd microphone between first microphone and second microphone is arranged on step in an array with first microphone, second microphone and the 3rd microphone.

5. according to the method for claim 2, wherein, response region is included in the zone before the resonance response.

6. according to the method for claim 2, wherein, response region is included in the zone after the resonance response.

7. according to the method for claim 2, wherein, response region is included in after the linear response region and the zone before resonance response.

Images