TWI787977B - Loudspeaker controller for estimating fundamental resonance frequency of loudspeaker and method for estimating fundamental resonance frequency of loudspeaker - Google Patents

Abstract

A loudspeaker controller for estimating a fundamental resonance frequency of a loudspeaker includes: an amplifier circuit, arranged to generate a driving signal of the loudspeaker according to an audio input signal; a sensing circuit, arranged to sense characteristics of the driving signal to generate a measurement signal; a plurality of band pass filter circuits, arranged to filter the measurement signal to generate a plurality of filter outputs, respectively, wherein the plurality of band pass filter circuits have different passbands; and an estimation circuit, arranged to estimate the fundamental resonance frequency according to the plurality of filter outputs.

Description

A loudspeaker controller for estimating the fundamental resonant frequency of a loudspeaker and for estimating Method for Measuring the Basic Resonant Frequency of Loudspeakers

This publication relates to frequency estimation, and in particular to a method for estimating a fundamental resonant frequency of a loudspeaker and related loudspeaker controllers.

A loudspeaker is a device with a voice coil that moves a diaphragm and converts electrical signals into sound signals, however, for input signals that cause larger diaphragm displacements, larger diaphragms Membrane displacement may damage the speaker. To avoid the above problems, an operating frequency of the speaker can be controlled within the fundamental resonance frequency of the speaker. In order to find the basic resonance frequency of the speaker, you can first obtain the impedance curve of the speaker, where the horizontal axis of the impedance curve is the frequency, and the vertical axis of the impedance curve is the impedance. Furthermore, the basic resonance frequency of the speaker It can be found from the impedance graph by identifying a frequency corresponding to the maximum value of the impedance curve. Note that the impedance curve of a speaker may change with temperature, therefore, the fundamental resonant frequency of a speaker is not a fixed value.

In order to obtain the impedance curve of the loudspeaker and the self-impedance curve to find the fundamental resonance frequency, a typical time-domain impedance measurement or a typical frequency-domain impedance measurement can be performed on the loudspeaker. The typical time-domain impedance measurement has Advantages of low cost and high accuracy, however, the typical time-domain impedance measurement cannot dynamically monitor the impedance and fundamental resonant frequency while the speaker is being driven for audio playback and requires a frequency sweep. For this typical frequency-domain impedance measurement, although the impedance and fundamental resonant frequency can be dynamically monitored and no frequency sweep is required while the loudspeaker is being driven for audio playback, the fast Fourier transform in the frequency-domain impedance measurement Transformation (fast fourier transformation, FFT) is quite complicated and may result in high hardware cost.

Therefore, an object of the present invention is to provide a method for estimating a fundamental resonance frequency of a loudspeaker. In addition, in order to achieve the advantages of low cost and high accuracy, the method can also be used for audio playback by the loudspeaker. The impedance (especially the fundamental resonant frequency) is dynamically monitored during driving.

According to an embodiment of the present invention, a method for estimating a fundamental resonance frequency of a loudspeaker is disclosed, the method may include: generating a driving signal of the loudspeaker according to an audio input signal; sensing the driving signal characteristics to generate a measurement signal; filter the measurement signal by a plurality of bandpass filters with different passbands to generate a plurality of filter outputs; and estimate the basic Resonance frequency.

In addition to the above method, the present invention further discloses a speaker controller, which may include: an amplifier circuit for generating a driving signal of the speaker according to an audio input signal; a sensing circuit for sensing Measuring the characteristics of the driving signal to generate a measurement signal; a plurality of bandpass filter circuits are used to respectively filter the measurement signal to generate a plurality of filter outputs, wherein the A plurality of bandpass filters have different passbands; and an estimation circuit is used for estimating the fundamental resonance frequency according to the outputs of the plurality of filters.

The present invention may have at least the following advantages/benefits. Compared with typical time-domain impedance measurements, the disclosed basic resonant frequency estimation scheme using a set of bandpass filter circuits with passbands centered at different frequencies can be used in loudspeakers for audio playback. Impedance (especially the fundamental resonant frequency) is dynamically monitored while being driven, and frequency sweeping is not required. Compared with typical frequency-domain impedance measurements, the disclosed basic resonant frequency estimation scheme using a set of bandpass filter circuits with passbands centered at different frequencies does not require complex fast Fourier transforms, And can be implemented with low hardware cost.

10:Speaker controller

12: Amplifier circuit

14: Sensing circuit

28_1~28_N: Bandpass filter circuit

30: Estimation circuit

50: speaker

A_IN: audio input signal

A_DRV: drive signal

S_M: Measurement signal

BPFOUT_1~BPFOUT_N: filter output

F _o : Fundamental resonance frequency

16: Current sensing circuit

18: Voltage sensing circuit

20: Preprocessing circuit

22: Low-pass filter circuit

23_1, 23_2: low pass filter

I(t): measurement current signal

I'(t): Low-pass filtered current signal

V(t): Measurement voltage signal

V’(t): Low-pass filtered voltage signal

24: Downsampling circuit

S_I: Reduce the sampling current signal

S_V: Reduced sampling voltage signal

29_11~29_N1, 29_12~29_N2: bandpass filter

BPFI_1~BPFI_N: bandpass filter current signal

BPFV_1~BPFV_N: Band-pass filtered voltage signal

32: Smoothing filter circuit

36_1~36_N: Alpha filter circuit

37_11~37_N1, 37_12~37_N2: Alpha filter

SFI_1~SFI_N: Smooth current signal

SFV_1~SFV_N: Smooth voltage signal

38: Processing circuit

40: Intensity Threshold Circuit

MAG: Strength

TH: Intensity Threshold

S80~S96, S98: steps

FIG. 1 is a block diagram of a speaker controller for estimating the fundamental resonance frequency of a speaker according to an embodiment of the present invention.

FIG. 2 is a schematic diagram of an implementation example of the speaker controller shown in FIG. 1 according to an embodiment of the present invention.

FIG. 3 is a flowchart of a method for estimating the fundamental resonance frequency of a speaker according to an embodiment of the present invention.

FIG. 4 is a schematic diagram of an impedance curve of a loudspeaker obtained by frequency-domain impedance measurement with fast Fourier transform.

FIG. 5 is a schematic diagram of estimating the fundamental resonance frequency of the loudspeaker by the method shown in FIG. 3 according to a first embodiment of the present invention.

FIG. 6 is a schematic diagram of another impedance curve of a loudspeaker obtained by frequency-domain impedance measurement with fast Fourier transform.

FIG. 7 is a schematic diagram of estimation of the fundamental resonant frequency of the loudspeaker by the method shown in FIG. 3 according to a second embodiment of the present invention.

FIG. 8 is a schematic diagram of yet another impedance curve of a loudspeaker obtained by frequency-domain impedance measurement with fast Fourier transform.

FIG. 9 is a schematic diagram of estimating the fundamental resonance frequency of the loudspeaker by the method shown in FIG. 3 according to a third embodiment of the present invention.

FIG. 10 is a schematic diagram of another implementation example of the speaker controller shown in FIG. 1 according to an embodiment of the present invention.

FIG. 11 is a flow chart of another method for estimating the fundamental resonant frequency of a loudspeaker according to an embodiment of the present invention.

FIG. 12 is a schematic diagram of yet another implementation example of the speaker controller shown in FIG. 1 according to an embodiment of the present invention.

FIG. 1 is a block diagram of a speaker controller 10 for estimating the fundamental resonant frequency F _o of a speaker 50 according to an embodiment of the present invention. As shown in FIG. 1, the speaker controller 10 is coupled to the speaker 50 and used to estimate the fundamental resonance frequency F _o of the speaker 50. It should be noted that the speaker 50 has the highest impedance at its fundamental resonance frequency F _o , Therefore, estimating the fundamental resonance frequency F _o of the loudspeaker 50 can be realized by estimating the highest impedance of the loudspeaker 50 . The speaker controller 10 may include an amplifier circuit 12, a sensing circuit 14, a plurality of band pass filter (band pass filter, BPF) circuits 28_1, 28_2, . . . BPF circuit") and an estimation circuit 30, wherein "N" may represent a positive integer greater than one (that is, N≧2). The amplifier circuit 12 is used for receiving an audio input signal A_IN and generating a driving signal A_DRV of the speaker 50 according to the audio input signal A_IN. The sensing circuit 14 is coupled to the amplifier circuit 12 and the speaker 50, and is used for sensing the characteristics of the driving signal A_DRV and generating a measurement signal S_M. The band-pass filter circuits 28_1~28_N are coupled to the sensing circuit 14, and are used to filter the measurement signal S_M and generate a plurality of filter outputs BPFOUT_1~BPFOUT_N respectively, wherein the band-pass filter circuits 28_1~28_N have different passbands (passband), therefore, when the same measurement signal S_M is input to the band-pass filter circuits 28_1-28_N, the filter outputs BPFOUT_1-BPFOUT_N may be different. The estimation circuit 30 is coupled to the bandpass filter circuits 28_1~28_N, and is used for estimating the fundamental resonant frequency F _o of the speaker 50 according to the filter outputs BPFOUT_1~BPFOUT_N. During being driven, the speaker controller 10 can estimate the fundamental resonant frequency F _o of the speaker 50 in a real-time manner.

Compared with the typical time-domain impedance measurement, the basic resonant frequency estimation scheme disclosed by the present invention using a set of bandpass filter circuits 28_1~28_N with passbands centered at different frequencies can be used in the loudspeaker 50 for Impedance (especially the fundamental resonant frequency) is dynamically monitored while being driven for audio playback, and frequency sweeping is not required.

Compared with typical frequency-domain impedance measurements, the basic resonant frequency estimation scheme disclosed in the present invention using a set of bandpass filter circuits 28_1~28_N with passbands centered on different frequencies does not require complicated fast Fourier transform, and can be implemented with low hardware cost. Further details of the basic resonance frequency estimation scheme disclosed in the present invention are described below with reference to the accompanying drawings.

FIG. 2 is a schematic diagram of an implementation example of the speaker controller shown in FIG. 1 according to an embodiment of the present invention. As mentioned above, the sensing circuit 14 is used to sense the characteristics of the driving signal A_DRV and generate the measurement signal S_M. For example, the characteristics of the driving signal A_DRV may include a voltage value and a current value, as shown in FIG. 2 The sensing circuit 14 may include a current sensing circuit 16 , a voltage sensing circuit 18 and a preprocessing circuit 20 . The current sensing circuit 16 can measure the current flowing through one of the voice coils of the loudspeaker 50 A current is used to generate a measurement current signal I(t). The voltage sensing circuit 18 can measure a voltage across the voice coil of the speaker 50 to generate a measurement voltage signal V(t). The preprocessing circuit 20 is used to The measurement signal S_M is generated according to the measurement current signal I(t) and the measurement voltage signal V(t).

In this embodiment, the preprocessing circuit 20 may include a low pass filter (LPF) circuit 22 (for brevity, labeled as "LPF circuit") and a downsampling circuit 24, wherein the downsampling circuit 24 is coupled to the low-pass filter circuit 22 . The low-pass filter circuit 22 may include a first low-pass filter 23_1 and a second filter 23_2 (for brevity, marked as "LPF ₁ " and "LPF ₂ " respectively), wherein the first low-pass filter The device 23_1 can receive the measured current signal I(t) generated by the current sensing circuit 16, and can low-pass filter the measured current signal I(t) to generate a low-pass filtered current signal I'(t), and the second The second filter 23_2 can receive the measured voltage signal V(t) generated by the voltage sensing circuit 18, and can low-pass filter the measured voltage signal V(t) to generate a low-pass filtered voltage signal V'(t). In order to reduce computational complexity and/or increase accuracy, the downsampling circuit 24 can receive the low-pass filtered current signal I'(t) and the low-pass filtered voltage signal V'(t), and downsample the low-pass filtered signal respectively. The current signal I'(t) and the low-pass filtered voltage signal V'(t) are used to generate a downsampled current signal S_I and a downsampled voltage signal S_V, wherein the measurement signal S_M shown in Figure 1 may include a second The downsampled current signal S_I and downsampled voltage signal S_V are shown in the figure.

According to this embodiment, the sensing circuit 14 can transmit the measurement signal S_M to a plurality of band-pass filter circuits 28_1~28_N, wherein the measurement signal S_M can include a current signal and a voltage signal (that is, the downsampled current signal S_I and lower the sampling voltage signal S_V). It should be noted that, depending on actual design considerations, the number of band-pass filter circuits 28_1~28_N implemented in the speaker controller 10, the frequency of each of the band-pass filter circuits 28_1~28_N wide (bandwidth) and/or the center frequency of the passband of each bandpass filter circuit in the bandpass filter circuits 28_1~28_N (also That is, the position of the pass band of each band-pass filter circuit in the band-pass filter circuits 28_1~28_N) can be adjusted, for example, the band-pass filter circuits 28_1~28_N can be provided by the manufacturer of the speaker 50 The nominal (nominal) fundamental resonant frequency of the loudspeaker 50 is configured to have respective passbands fixedly located/distributed in a frequency range, and for example, the bandpass filter circuits 28_1~28_N can be used for the loudspeaker 50 according to The time-varying fundamental resonant frequency of the loudspeaker 50 measured during audio playback while being driven is configured to have respective passbands dynamically located/distributed over a frequency range. In short, any loudspeaker controller that utilizes a set of bandpass filter circuits with passbands centered at different frequencies for fundamental resonant frequency estimation (or impedance estimation) falls within the scope of the present invention.

In addition, each of the band-pass filter circuits 28_1~28_N may include two band-pass filters. For example, the band-pass filter circuit 28_1 includes a first band-pass filter 29_11 and a first band-pass filter 29_11. The second band-pass filter 29_12 (labeled as "BPF ₁₁ " and "BPF ₁₂ " respectively for the sake of brevity); the band-pass filter circuit 28_2 includes a first band-pass filter 29_21 and a second band-pass filter device 29_22 (for brevity, respectively marked as "BPF ₂₁ " and "BPF ₂₂ "); and the bandpass filter circuit 28_N includes a first bandpass filter 29_N1 and a second bandpass filter 29_N2 (for For brevity, labeled "BPF _N1 " and "BPF _N2 " respectively). The first band-pass filter and the second band-pass filter of the same band-pass filter circuit have the same center frequency (that is, they are located at the same position), wherein the first band-pass filter can be used to receive signals from the sensing circuit 14. current signal (i.e. downsampled current signal S_I), and by filtering the current signal to generate a band-pass filtered current signal, the second band-pass filter can be used to receive the voltage signal from the sensing circuit 14 (i.e. downsampled voltage signal S_V ), and a band-pass filtered voltage signal is generated by filtering the voltage signal, and a filter output of the band-pass filter circuit includes a band-pass filtered current signal and a band-pass filtered voltage signal. For example, the filter output BPFOUT_1 includes the band-pass filtered current signal BPFI_1 and the band-pass filtered voltage signal BPFV_1, the filter output BPFOUT_2 includes the band-pass filtered current signal BPFI_2 and the band-pass filtered voltage signal BPFV_2, and the filter output BPFOUT_N includes There are band-pass filtered current signal BPFI_N and band-pass filtered voltage signal BPFV_N.

It should be noted that, for estimating the time-varying fundamental resonant frequency of the loudspeaker 50, before the fundamental resonant frequency estimation starts, the bandpass filter circuits 28_1˜28_N can be pre-placed according to the nominal fundamental resonant frequency of the loudspeaker 50 (pre -position) in a frequency range, however, this is only for illustrative purposes, and the present invention is not limited thereto.

The estimation circuit 30 of the speaker controller 10 can include a smoothing filter circuit 32 and a processing circuit 38. The smoothing filter circuit 32 can be used to receive the filter outputs BPFOUT_1~BPFOUT_N from the bandpass filter circuits 28_1~28_N, and by Smoothing the filter outputs BPFOUT_1~BPFOUT_N respectively to generate a plurality of smoothing filter outputs. In this embodiment, the smoothing filter circuit may include a plurality of alpha filter (alpha filter) circuits 36_1, 36_2, . The alpha filter circuits 36_1~36_N are respectively coupled to the bandpass filter circuits 28_1~28_N. In addition, each of the alpha filter circuits 36_1~36_N may include a first alpha filter and a second alpha filter (respectively labeled as "α filter ₁ " for brevity) and "α filter ₂ "), for example, the alpha filter circuit 36_1 includes a first filter 37_11 and a second alpha filter 37_12 (which are respectively coupled to the first bandpass filter 29_11 and the second band-pass filter 29_12), the alpha filter circuit 36_2 includes a first filter 37_21 and a second alpha filter 37_22 (which are respectively coupled to the first band-pass filter 29_21 and the second band-pass filter 29_22) , and the alpha filter circuit 36_N includes a first filter 37_N1 and a second alpha filter 37_N2 (which are respectively coupled to the first bandpass filter 29_N1 and the second bandpass filter 29_N2).

Since the filter output received by an alpha filter circuit includes a current signal and a voltage signal, the smoothing filter output generated by an alpha filter circuit includes a current signal and a voltage signal. For each smoothing filter output which includes a smoothed current signal and a smoothed voltage signal, the first alpha filter of the alpha filter circuit can be used with a bandpass filter circuit to receive the bandpass filtered current signal and generate the smoothed The current signal, and the second alpha filter of the alpha filter circuit can be used for the band-pass filter circuit to receive the band-pass filtered voltage signal and generate a smoothed voltage signal. As shown in FIG. 2, the smoothing filter output generated by the alpha filter circuit 36_1 includes a smoothed current signal SFI_1 and a smoothed voltage signal SFV_1, wherein the smoothed current signal SFI_1 is obtained by passing the band-pass filtered current signal BPFI_1 through the first alpha and the smoothed voltage signal SFV_1 is obtained by passing the band-pass filtered voltage signal BPF V_1 through the second alpha filter 37_12; the smoothing filter output generated by the alpha filter circuit 36_2 includes the smoothed current The signal SFI_2 and the smoothed voltage signal SFV_2, wherein the smoothed current signal SFI_2 is obtained by passing the band-pass filtered current signal BPFI_2 through the first alpha filter 37_21, and the smoothed voltage signal SFV_2 is obtained by passing the band-pass filtered voltage signal BPF V_2 through the first alpha filter 37_21 Two alpha filters 37_22 are obtained; and the smoothing filter output generated by the alpha filter circuit 36_N includes a smoothed current signal SFI_N and a smoothed voltage signal SFV_N, wherein the smoothed current signal SFI_N is obtained by passing the band-pass filtered current signal BPFI_N The first alpha filter 37_N1 is obtained, and the smoothed voltage signal SFV_N is obtained by passing the band-pass filtered voltage signal BPF V_N through the second alpha filter 37_N2 .

The alpha filter circuit in the smoothing filter circuit 32 can convert a filter output into a smoothing filter output to avoid or reduce the phase difference between the current signal and the voltage signal, that is, by the alpha filter The phase difference between the current signal and the voltage signal in the smoothing filter output produced by the circuit is smaller than the phase difference between the current signal and the voltage signal in the filter output sent to the alpha filter circuit, so that , the accuracy of fundamental resonance frequency estimation can be be improved.

For each smoothing filter output generated by the smoothing filter circuit 32 (especially the alpha filter circuits 36_1~36_N in the smoothing filter circuit 32), the processing circuit 38 can be used to divide the smoothed voltage signal by the smoothed current signal to generate an impedance value, in addition, the processing circuit 38 can also be used to estimate the speaker by comparing a plurality of impedance values {SFV_1/SFI_1, SFV_2/SFI_2, . . . , SFV_N/SFI_N} obtained from the output of the smoothing filter The fundamental resonant frequency F _o of 50.

In the case where a maximum is found among the self-impedance values {SFV_1/SFI_1, SFV_2/SFI_2, ..., SFV_N/SFI_N}, the fundamental resonance frequency F _o of the loudspeaker 50 is estimated as a derivation involving The center frequency of the bandpass filter circuit, for example, if the center frequency of the bandpass filter circuit corresponding to the maximum value among the impedance values {SFV_1/SFI_1, SFV_2/SFI_2, ..., SFV_N/SFI_N} The frequency is 200 Hz (hertz, Hz), and the fundamental resonance frequency F _o of the loudspeaker 50 can be estimated to be 200 Hz.

In another case where two impedance values with the same maximum value are found among the self-impedance values {SFV_1/SFI_1, SFV_2/SFI_2, ..., SFV_N/SFI_N}, the fundamental resonance frequency F _o of the loudspeaker 50 can be estimated Measured as an average of the center frequencies of the two bandpass filter circuits related to the derivation of the two impedance values having the same maximum value, for example, if corresponding to the impedance values {SFV_1/SFI_1, SFV_2/SFI_2, . .., SFV_N/SFI_N} The center frequency of the band-pass filter circuit of the maximum value is 200 Hz, and corresponds to The center frequency of another band-pass filter circuit with the same maximum value in is 210 Hz, then the fundamental resonant frequency F _o of loudspeaker 50 can be estimated to be 205 Hz, however, this is only for illustrative purposes, the present invention It is not limited to this. Alternatively, the fundamental resonance frequency F _o of the loudspeaker 50 can be estimated as any frequency value in a frequency range from 200 Hz to 210 Hz.

FIG. 3 is a flowchart of a method for estimating the fundamental resonance frequency of a speaker according to an embodiment of the present invention. If the same result can be obtained, the steps do not have to be executed sequentially according to the flow shown in Figure 3. For example, the method shown in Figure 3 can be implemented by the speaker controller 10 shown in Figure 2 accomplish.

In step S80 , a driving signal A_DRV of the speaker 50 is generated according to the audio input signal A_IN.

In step S82 , the current flowing through the voice coil of the speaker 50 is measured to generate a measured current signal I(t).

In step S84 , the voltage across the voice coil of the speaker 50 is measured to generate a measured voltage signal V(t).

In step S86, the measured current signal I(t) is low-pass filtered to generate a low-pass filtered current signal I'(t), and the measured voltage signal V(t) is low-pass filtered to generate a low-pass filtered voltage signal V' (t).

In step S88, the low-pass filtered current signal I'(t) and the low-pass filtered voltage signal V'(t) are respectively down-sampled to generate a down-sampled current signal S_I and a down-sampled voltage signal S_V.

In step S90, bandpass filter circuits 28_1~28_N with different passbands (for example, passbands with different center frequencies) are used to generate bandpass filtered current signals BPFI_1~BPFI_N and bandpass filter circuits BPFI_1~BPFI_N The pass-filtered voltage signals BPFV_1-BPFV_N, wherein the band-pass-filtered current signal and the band-pass-filtered voltage signal are generated by each of the band-pass filter circuits 28_1-28_N.

In step S92 , the band-pass filtered current signals BPFI_1 ˜ BPFI_N are flattened to generate smooth current signals SFI_1 ˜ SFI_N, and the band-pass filtered voltage signals BPFV_1 ˜ BPFV_N are flattened to generate smooth voltage signals SFV_1 ˜ SFV_N.

In step S94, a plurality of impedance values are generated according to the smoothed current signals SFI_1~SFI_N and the smoothed voltage signals SFV_1~SFV_N, wherein for each smoothing filter output including a smoothed current signal and a smoothed voltage signal, the smoothed voltage The signal is divided by the smoothed current signal to generate an impedance value.

In step S96 , the fundamental resonant frequency F _o of the speaker 50 is estimated according to the center frequency of one or more bandpass filter circuits corresponding to the maximum value of the plurality of impedance values.

Since those skilled in the art can easily understand the operation of each step shown in Fig. 3 through the contents of the instruction manual of the speaker controller 10 shown in Fig. 1 and Fig. 2, for the sake of brevity, similar contents It will not be repeated here.

In order to clarify that compared with the method of obtaining the fundamental resonance frequency by using fast Fourier transform analysis, the present invention can also estimate the fundamental resonance frequency with high accuracy and low cost. The following uses frequency domain impedance quantities with fast Fourier transform respectively The basic resonant frequency of a loudspeaker playing a set of specific music is analyzed and obtained through the measurement and the method of the present invention. Please refer to FIG. 4 and FIG. 5 together. FIG. 4 is a schematic diagram of the impedance curve of the loudspeaker 50 obtained by frequency-domain impedance measurement with fast Fourier transform. FIG. 5 is a schematic diagram of estimating the fundamental resonance frequency F _o of the loudspeaker 50 by the method shown in FIG. 3 according to a first embodiment of the present invention. As shown in FIG. 4 , by frequency-domain impedance measurement with FFT, it can be known that the fundamental resonant frequency F _o of the speaker 50 is approximately equal to 190 Hz. As shown in FIG. 5, by the method shown in FIG. 3, before estimating the fundamental resonant frequency F _o of the loudspeaker 50, the five bandpass filter circuits 28_1~28_5 (N=5) can be based on the nominal fundamental The resonant frequencies are preset at 200 Hz, 210 Hz, 220 Hz, 230 Hz, and 240 Hz, respectively, but the present invention is not limited thereto.

As shown in FIG. 5, the band-pass filter circuit (ie, band-pass filter circuit 28_1) at 200 Hz corresponds to the highest impedance value, wherein the highest impedance value is approximately equal to 19 ohms. Therefore, the fundamental resonance frequency of the loudspeaker 50 F _o can be estimated to the nearest 200 Hz. In the estimation of the fundamental resonant frequency of the loudspeaker, an error is usually within the tolerance range of less than 50 Hz, and the difference between 200 Hz and 190 Hz (that is, 10 Hz) is less than 50 Hz. Compared with the frequency-domain impedance measurement of , the method shown in FIG. 3 can estimate the fundamental resonant frequency F _o of the loudspeaker 50 with high accuracy and low cost.

Please refer to FIG. 6 and FIG. 7 together. FIG. 6 is a schematic diagram of another impedance curve of the loudspeaker 50 obtained by frequency-domain impedance measurement with fast Fourier transform. FIG. 7 is a schematic diagram of estimating the fundamental resonance frequency F _o of the loudspeaker 50 by the method shown in FIG. 3 according to a second embodiment of the present invention. As shown in FIG. 6, by frequency-domain impedance measurement with FFT, it can be known that the fundamental resonance frequency F _o of the speaker 50 is approximately equal to 95 Hz. As shown in FIG. 7, by the method shown in FIG. 3, although the fundamental resonance frequency F _o shown in _FIG . 6 is approximately equal to 95 Hz, five The bandpass filter circuits 28_1~28_5 (N=5) can still be pre-placed at 200 Hz, 210 Hz, 220 Hz, 230 Hz and 240 Hz, respectively, where these frequencies are far from the fundamental resonance frequency shown in FIG. 6 F _o , but the present invention is not limited thereto.

As shown in FIG. 7, the band-pass filter circuit (ie, band-pass filter circuit 28_1) at 200 Hz corresponds to the highest impedance value, wherein the highest impedance value is approximately equal to 14 ohms. Therefore, the fundamental resonance frequency of the loudspeaker 50 F _o can be estimated to the nearest 200 Hz. In the estimation of the fundamental resonant frequency of the loudspeaker, an error is usually within the tolerance range of less than 50 Hz, and the difference between 200 Hz and 95 Hz (that is, 105 Hz) is greater than 50 Hz, although by the third method of the present invention The method shown in the figure cannot use the central frequency setting of the band-pass filter circuit to accurately estimate the fundamental resonance frequency F _o of the speaker 50, because the frequency of the band-pass filter circuit is placed closer to the fundamental resonance frequency F of the speaker 50 _o , the impedance value corresponding to the bandpass filter circuit is larger, so the trend of the fundamental resonance frequency F _o of the loudspeaker 50 can still be known by the method shown in FIG. 3 . The current estimation result of the fundamental resonance frequency F _o of the loudspeaker 50 can be utilized as a reference for adaptively adjusting the center frequency setting of the bandpass filter circuit, so _that After properly adjusting the central frequency setting of the bandpass filter circuit, the fundamental resonance frequency F _o of the loudspeaker 50 can be accurately estimated by the method shown in FIG. 3 of the present invention.

Please refer to FIG. 8 and FIG. 9 together. FIG. 8 is a schematic diagram of yet another impedance curve of the loudspeaker 50 obtained by frequency-domain impedance measurement with fast Fourier transform. FIG. 9 is a schematic diagram of estimating the fundamental resonance frequency F _o of the loudspeaker 50 by the method shown in FIG. 3 according to a third embodiment of the present invention. As shown in FIG. 8, by frequency-domain impedance measurement with FFT, it can be known that the fundamental resonance frequency F _o of the speaker 50 is approximately equal to 205 Hz. As shown in FIG. 9, by the method shown in FIG. 3, before estimating the fundamental resonant frequency F _o of the loudspeaker 50, the five bandpass filter circuits 28_1~28_5 (N=5) can be based on the nominal fundamental The resonant frequencies are preset at 200 Hz, 210 Hz, 220 Hz, 230 Hz, and 240 Hz, respectively, but the present invention is not limited thereto.

As shown in FIG. 9, the impedance value obtained from the filter output of the band-pass filter circuit (for example, band-pass filter circuit 28_1) at 200 Hz is the same as that obtained from the filter output of the band-pass filter circuit (for example, band-pass filter circuit 28_1) at 210 Hz. The impedance values obtained by the filter outputs of the filter circuit 28_2) are quite close, so it is quite difficult to determine which bandpass filter circuit has the highest impedance value. In this case, the fundamental resonant frequency F _o of the loudspeaker 50 can be estimated as an intermediate frequency between 200 Hz and 210 Hz (i.e. 205 Hz). In the estimation of the fundamental resonant frequency of the loudspeaker, an error usually is in the tolerance range of less than 50 Hz, and the estimated result of this embodiment is exactly the same as the fundamental resonance frequency of the loudspeaker shown in Fig. The method shown in FIG. 3 can estimate the fundamental resonant frequency F _o of the loudspeaker 50 with high accuracy and low cost.

It should be noted that when the speaker is driven at a low volume for audio playback, the estimation result of the method shown in Fig. 3 may have errors, for example, in Fig. In the time from second to the 15th second, the band-pass filter circuit at 200 Hz (that is, the band-pass filter circuit 28_1) corresponds to the highest impedance value, which is different from the band-pass filter circuit at 200 Hz (such as the band-pass filter circuit 28_1). The impedance value obtained from the filter output of the filter circuit 28_1) is not the same as the above case in which the impedance value obtained from the filter output of the band-pass filter circuit at 210 Hz (for example, the band-pass filter circuit 28_2) is quite close . To solve this problem, an intensity threshold can be added to the speaker controller to avoid errors when the speaker is driven at a low volume for audio playback.

FIG. 10 is a schematic diagram of another implementation example of the speaker controller 10 shown in FIG. 1 according to an embodiment of the present invention. In order to avoid errors when the loudspeaker 50 is driven at a low volume for audio playback, the estimation circuit 30 shown in FIG. 10 further includes an intensity threshold circuit 40, which is coupled to the amplifier circuit 12, The smoothing filter circuit 32 and the processing circuit 38 are used for comparing a magnitude MAG of the driving signal A_DRV with a magnitude threshold TH. When the intensity MAG of the driving signal A_DRV exceeds the intensity threshold TH, the estimation circuit 30 estimates the fundamental resonance frequency F _o according to the filter outputs of the bandpass filter circuits 28_1~28_N, for example, when the intensity MAG of the driving signal A_DRV When the intensity threshold TH is exceeded, the processing circuit 38 estimates the fundamental resonant frequency F _o by comparing the impedance values obtained from the smoothed current signals SFI_1~SFI_N and the smoothed voltage signals SFV_1~SFV_N, wherein the smoothed current signals SFI_1~SFI_N are The smoothed voltage signals SFV_1 ˜ SFV_N are obtained by filtering the band-pass filtered current signals BPFI_1 ˜ BPFI_N, and the smoothed voltage signals SFV_1 ˜ SFV_N are obtained by filtering the band-pass filtered voltage signals BPFV_1 ˜ BPFV_N. When the intensity MAG of the driving signal A_DRV does not exceed the intensity threshold TH, the estimation circuit 30 will not estimate the fundamental resonant frequency F _o according to the filter outputs of the bandpass filter circuits 28_1~28_N. Estimation error at low volume.

FIG. 11 is a flow chart of another method for estimating the fundamental resonant frequency of a loudspeaker according to an embodiment of the present invention. If the same result can be obtained, the steps do not have to be executed sequentially according to the flow shown in Figure 11. For example, the method shown in Figure 11 can be implemented by the speaker controller 10 shown in Figure 10 accomplish. The difference between the method shown in FIG. 3 and the method shown in FIG. 11 is that the method shown in FIG. 11 further includes step S98. In step S98, when the intensity MAG of the driving signal A_DRV exceeds the intensity threshold TH , the process enters step S96; when the intensity MAG of the driving signal A_DRV does not exceed the intensity threshold TH, the process returns to step S80.

In some embodiments of the present invention, only when the current intensity or a voltage intensity of the driving signal A_DRV is greater than the intensity threshold TH, the sensing circuit 14 can sense the characteristics of the driving signal A_DRV to generate the measurement signal S_M (the It may include measuring the current signal I(t) and measuring the voltage signal V(t)). For example, the intensity threshold TH can be set as a current threshold, and only when the current intensity of the driving signal is greater than the intensity threshold TH, The sensing circuit 14 is allowed to sense the characteristics of the driving signal A_DRV to generate the measurement signal No. S_M, but the present invention is not limited thereto.

FIG. 12 is a schematic diagram of yet another implementation example of the speaker controller 10 shown in FIG. 1 according to an embodiment of the present invention. In this embodiment, the voltage can be set to a fixed value, and the relationship between the current and the impedance is an inverse number. Therefore, as shown in FIG. 12, the sensing circuit 14 can include a current sensing circuit 16 and a pre-processing The circuit 20, the current sensing circuit 16 can measure a current flowing through the voice coil of the loudspeaker 50 to generate a measured current signal I(t), and the preprocessing circuit 20 is used to measure the current signal I(t) according to the measured current signal I(t) Generate measurement signal S_M. In this embodiment, the pre-processing circuit 20 may include a low-pass filter circuit 22 and a down-sampling circuit 24 , wherein the down-sampling circuit 24 is coupled to the low-pass filter circuit 22 . The low-pass filter circuit 22 can include a low-pass filter 23_1 (marked as “LPF ₁ ” for brevity), wherein the low-pass filter 23_1 can receive the measurement current signal I ( t), and the measured current signal I(t) can be low-pass filtered to generate a low-pass filtered current signal I'(t). In order to reduce computational complexity and/or increase accuracy, the downsampling circuit 24 can receive the low-pass filtered current signal I'(t) to generate a downsampled current signal S_I, wherein the measurement signal S_M shown in FIG. 1 can include Figure 12 shows the downsampled current signal S_I.

The sensing circuit 14 can transmit the measurement signal S_M to the bandpass filter circuits 28_1~28_N, wherein the measurement signal S_M can include a current signal (that is, reduce the sampling current signal S_I). In addition, the bandpass filter circuits 28_1~28_N Each band-pass filter circuit in 28_N may comprise a band-pass filter, for example, band-pass filter circuit 28_1 comprises a band-pass filter 29_11 (labeled "BPF ₁₁ " for brevity); The pass filter circuit 28_2 includes a band pass filter 29_21 (labeled "BPF ₂₁ " for brevity); and the band pass filter circuit 28_N includes a band pass filter 29_N1 (labeled "BPF 21" for brevity). _N1 "). The band-pass filter in each of the band-pass filter circuits 28_1~28_N can be used to receive the current signal from the sensing circuit 14 (for example, down-sample the current signal S_I), and filter the current signal to A band-pass filtered current signal is generated, and a filter output of the band-pass filter circuit includes the band-pass filtered current signal. For example, the filter output BPFOUT_1 includes the band-pass filtered current signal BPFI_1, the filter output BPFOUT_2 includes the band-pass filtered current signal BPFI_2, and the filter output BPFOUT_N includes the band-pass filtered current signal BPFI_N.

The estimation circuit 30 of the speaker controller 10 can include a smoothing filter circuit 32 and a processing circuit 38. The smoothing filter circuit 32 can be used to receive filter outputs from the pass filter circuits 28_1~28_N and smooth the filters by respectively output to produce a complex number of smoothing filter outputs. In this embodiment, the smoothing filter circuit 32 may include a plurality of alpha filter circuits 36_1, 36_2, ..., 36_N, wherein the alpha filter circuits 36_1~36_N are respectively coupled to the bandpass filter circuit 28_1 ~28_N. In addition, each of the alpha filter circuits 36_1~36_N may include an alpha filter (marked as “α filter” for brevity), for example, the alpha filter circuit 36_1 Including an alpha filter 37_11 (which is coupled to the band-pass filter 29_11), the alpha filter circuit 36_2 comprises an alpha filter 37_21 (which is coupled to the band-pass filter 29_21), and an alpha filter circuit 36_N includes an alpha filter 37_N1 (which is coupled to the bandpass filter 29_N1).

As shown in FIG. 12, a smoothing filter output generated by the alpha filter circuit 36_1 includes a smoothed current signal SFI_1, wherein the smoothed current signal SFI_1 is obtained by passing the band-pass filtered current signal BPFI_1 through the alpha filter 37_11 ; A smoothing filter output generated by the alpha filter circuit 36_2 includes a smooth current signal SFI_2, wherein the smooth current signal SFI_2 is obtained by passing the band-pass filtered current signal BPFI_2 through the alpha filter 37_21; and the alpha filter A smoothing filter output generated by the circuit 36_N includes the smoothed current signal SFI_N, wherein the smoothed current signal SFI_N is obtained by passing the band-pass filtered current signal BPFI_N through the alpha filter 37_N1.

In this embodiment, since the voltage is set to a fixed value, each smoothing filter output generated by the smoothing filter circuit 32 (especially the alpha filter circuits 36_1~36_N in the smoothing filter circuit 32) (especially the smooth current signal), the processing circuit 38 can be used to compare the strength of each smooth current signal, wherein when the strength of the smooth current signal is smaller, the basic resonance of the loudspeaker 50 estimated by the processing circuit 38 The larger the frequency F _o is, therefore, the fundamental resonance frequency F _o of the loudspeaker 50 is estimated in the case where a minimum is found among the intensities from the complex number of smooth current signals {SFI_1, SFI_2, ..., SFI_N} is the center frequency of the bandpass filter circuit involved in the derivation of the minimum, for example, if the band corresponding to the minimum among the intensities of the plurality of smoothed current signals {SFI_1, SFI_2, ..., SFI_N} Since the center frequency of the pass filter circuit is 200 Hz, the fundamental resonant frequency F _o of the loudspeaker 50 can be estimated to be 200 Hz. For the sake of brevity, similar content in this embodiment is not repeated here.

The above descriptions are only preferred embodiments of the present invention, and all equivalent changes and modifications made according to the scope of the patent application of the present invention shall fall within the scope of the present invention.

10:Speaker controller

12: Amplifier circuit

14: Sensing circuit

28_1~28_N: Bandpass filter circuit

30: Estimation circuit

50: speaker

A_IN: audio input signal

A_DRV: drive signal

S_M: Measurement signal

BPFOUT_1~BPFOUT_N: filter output

F _o : Fundamental resonance frequency

Claims (16)

A loudspeaker controller for estimating a fundamental resonance frequency of a loudspeaker, comprising: an amplifier circuit for generating a driving signal of the loudspeaker according to an audio input signal; a sensing circuit for sensing the The characteristics of the driving signal are used to generate a measurement signal; a plurality of bandpass filter circuits are used to respectively filter the measurement signal to generate a plurality of filter outputs, wherein the plurality of bandpass filters have different passbands; and a The estimation circuit is used for comparing an intensity of the driving signal with an intensity threshold, and estimating the fundamental resonance frequency according to the plurality of filter outputs in response to the intensity of the driving signal exceeding the intensity threshold. The speaker controller as described in claim 1, wherein the sensing circuit includes: a current sensing circuit for generating a measurement by measuring a current flowing through a voice coil of the speaker a current signal; a voltage sensing circuit for generating a measurement voltage signal by measuring a voltage across the voice coil of the loudspeaker; and a preprocessing circuit for generating a measurement voltage signal based on the measurement current signal and the quantity The voltage measurement signal is used to generate the measurement signal. The loudspeaker controller described in claim 2 of the patent application, wherein the preprocessing circuit includes: a low-pass filter circuit for filtering the measured current signal to generate a low-pass filtered current signal, and filtering the quantity measuring the voltage signal to generate a low-pass filtered voltage signal; and a downsampling circuit for downsampling the low-pass filtered current signal to generate a downsampled current signal, and downsampling the low-pass filtered voltage signal to generate a downsampled voltage signal; wherein the quantity output from the preprocessing circuit The test signal includes the downsampled current signal and the downsampled voltage signal. The loudspeaker controller as described in item 1 of the scope of the patent application, wherein the measurement signal includes a current signal and a voltage signal; each band-pass filter circuit of the plurality of band-pass filter circuits includes a first band-pass filter circuit filter and a second band-pass filter; for each band-pass filter circuit of the plurality of band-pass filter circuits, the first band-pass filter is used to generate a band-pass filter by filtering the current signal filtering the current signal, and the second bandpass filter for generating a bandpass filtered voltage signal by filtering the voltage signal; and a filter for each of the plurality of bandpass filter circuits The output includes the band-pass filtered current signal and the band-pass filtered voltage signal. The loudspeaker controller as described in item 4 of the scope of the patent application, wherein the estimation circuit includes: a smoothing filter circuit, which is used to level the plurality of filter outputs respectively to generate a plurality of smoothing filter outputs, wherein the Each smoothing filter output of the plurality of smoothing filter outputs includes a smoothed current signal and a smoothed voltage signal; and a processing circuit for estimating the fundamental resonance frequency according to the plurality of smoothing filter outputs. The loudspeaker controller as described in item 5 of the scope of patent application, wherein the smoothing filter circuit includes a plurality of alpha filter circuits; each alpha of the plurality of alpha filter circuits The alpha filter circuit includes a first alpha filter and a second alpha filter; and for each smoothing filter output of the plurality of smoothing filter outputs, the smoothed current signal is obtained by using the first alpha filter An alpha filter is generated, and the smoothed voltage signal is generated by using the second alpha filter. The speaker controller described in claim 5, wherein for each smoothing filter output of the plurality of smoothing filter outputs, the processing circuit is used to divide the smoothed voltage signal by the smoothed current signal to obtain An impedance value is generated; and the processing circuit is further used to estimate the fundamental resonant frequency by comparing a plurality of impedance values obtained from a plurality of smoothing filter outputs. The speaker controller as described in claim 1 of the claimed invention, wherein during the period when the speaker is driven for audio playback, the speaker controller estimates the fundamental resonance frequency of the speaker in real time. A method for estimating a fundamental resonance frequency of a loudspeaker, comprising: generating a driving signal of the loudspeaker according to an audio input signal; sensing a characteristic of the driving signal to generate a measurement signal; by having different filtering the measurement signal with a plurality of passband filters to generate a plurality of filter outputs; comparing an intensity of the driving signal with an intensity threshold; and responding to the intensity of the driving signal exceeding the intensity threshold , estimating the fundamental resonance frequency according to the plurality of filter outputs. The method described in claim 9, wherein the characteristic of the drive signal is sensed to The step of generating the measurement signal includes: generating a measurement current signal by measuring a current flowing through a voice coil of the loudspeaker; generating a measurement current signal by measuring a voltage across the voice coil of the loudspeaker measuring a voltage signal; and generating the measuring signal according to the measuring current signal and the measuring voltage signal. The method described in claim 10, wherein the step of generating the measurement signal according to the measurement current signal and the measurement voltage signal includes: low-pass filtering the measurement current signal to generate a low-pass filter current signal; low pass filtering the measured voltage signal to generate a low pass filtered voltage signal; downsampling the low pass filtered current signal to generate a downsampled current signal; and downsampling the low pass filtered voltage signal to generate a downsampled A voltage signal; wherein the measurement signal includes the downsampled current signal and the downsampled voltage signal. The method described in item 9 of the claimed scope of patents, wherein the measurement signal includes a current signal and a voltage signal, and each band-pass filter circuit of the plurality of band-pass filter circuits includes a first band-pass filter and a second band-pass filter, and the step of filtering the measurement signal through the plurality of band-pass filters with different passbands to generate the outputs of the plurality of filters includes: for the plurality of band-pass filters Each of the bandpass filter circuits of the filter circuit: filters the current signal by the first bandpass filter to generate a bandpass filtered current signal; and filters the voltage by the second bandpass filter signal to generate a band-pass filtered voltage signal; wherein a filter output of each of the plurality of band-pass filter circuits includes the band-pass filtered current signal and the band-pass filtered voltage signal. The method described in claim 12, wherein the step of estimating the fundamental resonance frequency according to the plurality of filter outputs includes: leveling the plurality of filter outputs respectively to generate a plurality of smoothing filter outputs, wherein Each smoothing filter output of the plurality of smoothing filter outputs includes a smoothed current signal and a smoothed voltage signal; and the fundamental resonance frequency is estimated according to the plurality of smoothing filter outputs. The method according to claim 13, wherein for each smoothing filter output of the plurality of smoothing filter outputs, the smoothed current signal is generated by using a first alpha filter, and the The smoothed voltage signal is generated by using a second alpha filter. The method described in claim 13, wherein the step of estimating the fundamental resonance frequency according to the plurality of smoothing filter outputs includes: for each smoothing filter output of the plurality of smoothing filter outputs, The smoothed voltage signal is divided by the smoothed current signal to generate an impedance value; and the fundamental resonant frequency is estimated by comparing a plurality of impedance values obtained from a plurality of smoothing filter outputs. The method according to claim 9, wherein the method estimates the fundamental resonant frequency of the speaker in a real-time manner while the speaker is driven for audio playback.

Images