Method and system for estimating loudspeaker voice coil offset based on signal envelope and medium

Abstract

The invention discloses a method and system for estimating loudspeaker voice coil offset based on signal envelope and a medium. According to the technical scheme, the method comprises the steps of obtaining a measurement signal of a loudspeaker to be measured, calculating an upper envelope line and a lower envelope line of the signal, calculating a signal envelope bias based on the envelope lines, dividing the envelope bias into a low frequency band and a high frequency band by taking the resonant frequency of the loudspeaker as a boundary, and calculating a signal envelope gradient by analyzing characteristic values of the bias in the two frequency bands. And finally, converting the gradient into a voice coil offset through polynomial transformation. The method is suitable for various signal types such as sound pressure, acceleration and displacement, normalization processing is carried out by introducing the reference loudspeaker, and the detection accuracy and the environment robustness are effectively improved. The device and the method have the beneficial effects of adapting to high-speed production lines and high-precision detection requirements.

Description

Technical Field

[0001] This invention relates to the field of electroacoustic device testing technology, and in particular to a method, system, and medium for estimating loudspeaker voice coil offset based on signal envelope. Background Technology

[0002] A loudspeaker's performance is highly dependent on the precise centering of its voice coil within the magnetic gap. Ideally, the voice coil should be centered within the gap. However, during mass production, due to process tolerances and assembly errors, the voice coil may deviate from its center position, resulting in what is known as "voice coil offset." Excessive voice coil offset can cause non-linear distortion in the loudspeaker, reducing audio quality and affecting product performance. Therefore, it is crucial to quickly and accurately detect and reject products with excessive voice coil offset during production quality inspection.

[0003] In existing technologies, methods for detecting voice coil offset are mainly divided into two categories: one is the traditional method, which relies on complex laser measurement or precision mechanical probes. Although these methods have high accuracy, they suffer from drawbacks such as expensive equipment, slow measurement speed, and difficulty in integration into high-speed production lines; the other is the electrical measurement method, which indirectly determines the offset by analyzing the loudspeaker's response signal. However, these methods are sensitive to the measurement environment (such as temperature, humidity, and sensor position) and generally lack universality for various signal types (such as sound pressure, acceleration, and displacement), resulting in poor robustness and inaccurate estimation results.

[0004] In summary, existing technologies have inherent contradictions in terms of detection efficiency, cost, robustness, and signal universality, making it difficult to meet the technical requirements of high-speed production lines for fast, stable, low-cost, and high-precision detection. Summary of the Invention

[0005] The main objective of this invention is to propose a method, system, and medium for estimating loudspeaker voice coil offset based on signal envelope. This aims to solve the technical problem that existing technologies suffer from inherent contradictions in terms of detection efficiency, cost, robustness, and signal universality, making it difficult to meet the requirements of high-speed production lines for fast, stable, low-cost, and high-precision detection.

[0006] To achieve the above objectives, this invention proposes a method for estimating loudspeaker voice coil offset based on signal envelope, comprising the following steps:

[0007] Acquire the measurement signal of the speaker under test;

[0008] Calculate the upper and lower envelopes of the measured signal;

[0009] Based on the upper and lower envelopes, the signal envelope bias is calculated;

[0010] Calculate the signal envelope tilt based on the aforementioned signal envelope bias;

[0011] The signal envelope tilt is converted into the voice coil offset of the loudspeaker under test through a polynomial transformation.

[0012] Furthermore, after calculating the signal envelope bias and before calculating the signal envelope tilt, the method further includes the following steps:

[0013] Obtain the reference signal envelope bias of the reference loudspeaker;

[0014] The normalized signal envelope bias is obtained by subtracting the signal envelope bias of the loudspeaker under test from the reference signal envelope bias.

[0015] The step of calculating the signal envelope tilt is performed based on the normalized signal envelope bias.

[0016] Furthermore, the step of calculating the signal envelope tilt includes:

[0017] Using the resonant frequency of the loudspeaker as the boundary, the signal envelope bias or normalized signal envelope bias is divided into two parts: a low-frequency band and a high-frequency band.

[0018] Calculate the algebraic sum of the envelope bias in the low-frequency band and the high-frequency band respectively, and select the maximum or minimum value of the envelope bias in the low-frequency band and the high-frequency band respectively according to the sign of the algebraic sum.

[0019] The signal envelope tilt is calculated using eigenvalues. The signal envelope tilt is the algebraic difference between low-frequency and high-frequency eigenvalues. The formula for calculating the signal envelope tilt is:

[0020] x=feature low -feature high

[0021] Where x is the signal envelope tilt, and is a scalar value; feature low Low-frequency feature values; high These are high-frequency characteristic values.

[0022] Furthermore, the polynomial transformation takes the form of:

[0023] y=k0+k1x+k2x 2 +…+k N x N

[0024] Where y is the voice coil offset; x is the signal envelope tilt, and is a scalar value; k0, k1, ..., k N is the polynomial coefficient; N is the polynomial order.

[0025] Furthermore, the polynomial coefficients are determined through a training process, which includes:

[0026] Acquire measurement signals from multiple loudspeaker samples with known voice coil offsets;

[0027] The corresponding signal envelope tilt is calculated based on the measured signal.

[0028] By using a regression algorithm, a polynomial relationship between the known voice coil offset and the signal envelope tilt is fitted, thereby determining the polynomial coefficients.

[0029] Furthermore, the loudspeaker sample with a known voice coil offset is obtained through at least one of the following methods:

[0030] Computer-based physical simulation of loudspeakers with different voice coil offsets;

[0031] Manufacturing physical speaker samples with different voice coil offsets;

[0032] Collect speaker data during production trial operation;

[0033] Different DC signals of varying magnitudes were applied to standard loudspeaker samples to simulate different voice coil offsets.

[0034] Furthermore, the measurement signal is a time-domain signal, and its type includes pressure signal, acceleration signal, or displacement signal.

[0035] Furthermore, after obtaining the voice coil offset, the method further includes the following steps:

[0036] The voice coil offset is compared with a preset tolerance range;

[0037] Based on the comparison results, the conclusion that the speaker under test is qualified or unqualified is output.

[0038] The present invention also proposes a loudspeaker production quality control system, including a measurement system and a processor. The measurement system is used to acquire measurement signals of the loudspeaker under test, and the processor operates using the method described above.

[0039] The present invention also proposes a computer-readable storage medium comprising a stored computer program, wherein the computer program, when running, controls the device or apparatus containing the computer-readable storage medium to perform the aforementioned method.

[0040] The technical solution of this invention includes acquiring the measurement signal of the loudspeaker under test, calculating the upper and lower envelopes of the signal, calculating the signal envelope bias based on the envelopes, dividing the envelope bias into low-frequency and high-frequency bands with the loudspeaker resonant frequency as the boundary, calculating the signal envelope tilt by analyzing the characteristic values ​​of the bias within the two frequency bands, and finally converting the tilt into a voice coil offset through polynomial transformation. This invention is applicable to various signal types such as sound pressure, acceleration, and displacement. By introducing a reference loudspeaker for normalization processing, it effectively improves the accuracy and environmental robustness of the detection. This invention has the beneficial effect of adapting to the requirements of high-speed production lines and high-precision testing. Attached Figure Description

[0041] Figure 1 This is a flowchart of the present invention;

[0042] Figure 2 This is a schematic diagram of the measurement signal and its upper and lower envelopes in an embodiment of the present invention;

[0043] Figure 3 Based on Figure 2 A comparative diagram of the calculated upper and lower envelope amplitudes;

[0044] Figure 4 The signal flow graph for calculating the signal envelope bias step;

[0045] Figure 5 Based on Figure 4 A schematic diagram of the calculated signal envelope bias waveform;

[0046] Figure 6 The signal flow graph for the normalized signal envelope biasing step;

[0047] Figure 7 A comparative diagram showing the effects of normalization processing;

[0048] Figure 8 A flowchart for the signal envelope tilt;

[0049] Figure 9 A block diagram of the quality control system for loudspeaker production. Detailed Implementation

[0050] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0051] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, top, bottom, inside, outside, vertical, horizontal, longitudinal, counterclockwise, clockwise, circumferential, radial, axial, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0052] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0053] This invention proposes a method for estimating loudspeaker voice coil offset based on signal envelope.

[0054] In embodiments of the present invention, such as Figures 1 to 8 As shown, the overall process of this method for estimating loudspeaker voice coil offset based on signal envelope is as follows: Figure 1 The block diagram shown is defined and integrated as a module. Figure 9 The measurement system shown operates as described. It should be noted that this technology is not a standalone software system, but rather exists as a module or plug-in to other software systems. Therefore, it can operate without a dedicated user interface. This technology relies on analyzing signals acquired by other measurement systems. The measurement system or solution is used to acquire and analyze loudspeaker measurement data. The voice coil offset detector operates as an auxiliary module within this measurement system. The measurement system transmits the acquired measurement signals to the voice coil offset detector module. The voice coil offset detector module analyzes the received acquired signals, calculates the voice coil offset, and determines whether the performance of the loudspeaker under test is acceptable. The detection results are fed back to the measurement system and can be included in the measurement report generated by the system. Furthermore, the voice coil offset detector can generate its own reports independently of the measurement system. For example, the detector can write the derived results to a database or text result file as needed.

[0055] This invention includes the following steps:

[0056] S100: Acquire measurement signal.

[0057] Specifically, the measurement system controls the speaker under test (DUT) to play a test signal (such as a sweep frequency signal) and simultaneously acquires its response signal. The measurement signal can be a time-domain signal such as sound pressure level, acceleration, or displacement. An exemplary test signal waveform is shown below. Figure 2 As shown by the blue curve.

[0058] S200: Calculate the signal envelope.

[0059] Specifically, the upper envelope env of the measured signal is calculated. upper and lower envelope env lower In this embodiment, the Hilbert transform method or similar techniques can be used. As a preferred implementation, the specific steps of the Hilbert transform method are: performing a Hilbert transform on the measured signal to obtain its analytic signal; the amplitude of the analytic signal is the upper envelope env. upper The negative number of the upper envelope is the lower envelope env. lower It is understood that in other embodiments, methods such as spline interpolation based on local maxima and minima can also be used to calculate the upper and lower envelopes.

[0060] like Figure 2 As shown, the two red curves are the calculated upper and lower envelopes, which track the evolution of the positive and negative amplitudes of the signal, respectively. Figure 3 This more clearly shows the comparison between the amplitudes of the upper and lower envelopes.

[0061] S300: Calculate envelope bias.

[0062] Specifically, such as Figure 4 As shown, the upper envelope and lower envelope are added together, and the result is obtained according to the formula bias=env. upper +env lower Calculate the signal envelope bias.

[0063] The symbols are annotated as follows:

[0064] bias: signal envelope bias.

[0065] env upper Upper envelope.

[0066] env lower : Lower envelope.

[0067] Since the lower envelope value is negative, this operation essentially calculates the difference between the upper and lower envelopes. This bias directly reflects the asymmetry of the signal waveform, and its calculation result is shown in the example below. Figure 5 As shown.

[0068] S400 (optional): Normalized envelope bias.

[0069] This is a preferred step to improve robustness. Specifically, such as... Figure 6 As shown, the reference envelope bias of the reference loudspeaker is first obtained. REF The reference loudspeaker is a stable standard sample with extremely small voice coil offset (e.g., less than 0.05 mm) confirmed by precise measurements, and its envelope bias is... REF Measurements should be pre-measured and stored in a stable measurement environment before biasing is performed. norm =bias DUT -bias REF The normalized envelope bias is obtained. norm .

[0070] The symbols are annotated as follows:

[0071] bias norm : Normalized envelope bias.

[0072] bias DUT : Envelope bias of the speaker under test.

[0073] bias REF : The envelope bias of the reference speaker.

[0074] Figure 7 The upper part shows a comparison of the envelope bias of the tested and reference loudspeakers, while the lower part shows the normalized results, which effectively compensate for environmental variations.

[0075] S500: Calculate the envelope tilt.

[0076] Specifically, such as Figure 8 As shown, this step includes:

[0077] Frequency band segmentation: The (normalized) envelope bias is divided into low-frequency and high-frequency bands, using the resonant frequency of the loudspeaker as the boundary. The resonant frequency can be a fixed known parameter of this loudspeaker model, obtained by pre-measuring its impedance curve; or it can be estimated in real time from the spectrum of the currently measured signal by finding the frequency corresponding to the point of maximum amplitude.

[0078] Feature extraction: Calculate the algebraic sum (i.e., the sum including the sign) of all offset values ​​within each of the two frequency bands. Then, based on the sign of the algebraic sum, select the maximum (when the algebraic sum is positive) or minimum (when the algebraic sum is negative) offset value within each frequency band as the feature value. It should be noted that if the sum of a segment is positive, the maximum value of that segment is used; if the sum of a segment is negative, the minimum value of that segment is used.

[0079] Eigenvalue Envelope Tilting Calculation: The signal envelope tilt is the algebraic difference between low-frequency and high-frequency eigenvalues. The formula for calculating the signal envelope tilt is as follows:

[0080] x=feature low -feature high

[0081] Where x is the signal envelope tilt, a scalar value, and is related to the speaker voice coil offset; feature low Low-frequency feature values; high These are high-frequency characteristic values.

[0082] It is understandable that a complex nonlinear relationship exists between the signal envelope tilt and the physical voice coil offset. This invention employs a polynomial model to characterize this relationship based on the following facts: First, the model's adaptability; the polynomial coefficients and order N can be tailored to different product types through a training process, allowing the same method to be flexibly applied to loudspeaker products with different acoustic characteristics. Second, ease of implementation: polynomial transformation mainly involves basic arithmetic operations, with a low computational burden, facilitating implementation on various computing platforms, including low-power microcontrollers, meeting the production line's requirements for detection speed. Therefore, the polynomial model provides a practical and effective solution for achieving high-precision, high-efficiency voice coil offset estimation. The polynomial coefficients can be obtained through various training methods, such as computer simulation, creating physical samples, using production data, or applying a DC offset.

[0083] Specifically, a process for determining polynomial coefficients is as follows: This process is typically performed during production line setup or product model changes, and includes the following steps:

[0084] Obtain a training sample set. Acquire multiple loudspeakers with known voice coil offsets as training samples. These samples can be obtained through at least one of the following methods: performing computer-based physical simulations of loudspeakers with different voice coil offsets; manufacturing physical loudspeaker samples with different voice coil offsets; collecting loudspeaker data during production trial runs and measuring their offsets using benchmark methods; or applying DC signals of different magnitudes to standard loudspeaker samples to simulate different voice coil offsets.

[0085] Feature extraction. For each training sample, the measurement signal is acquired according to steps S100 to S500 of this invention, and the corresponding signal envelope tilt (denoted as xi) is calculated.

[0086] Establish a mapping relationship. For each training sample, the known voice coil offset (denoted as yi) and the calculated signal envelope tilt (denoted as xi) form a data pair. Using a regression algorithm (e.g., least squares), perform curve fitting on this dataset to find a polynomial function y=f(x) that best fits all data points (xi, yi). This fitting process determines the optimal coefficients k0, k1, k2, ..., k of the polynomial. N And the order N.

[0087] Deploy the model. The optimal polynomial coefficients and order obtained from training are fixed in a configuration file or directly written into the processor's program for online real-time voice coil offset calculation of subsequent speakers under test (DUTs).

[0088] S600: Convert to voice coil offset.

[0089] Specifically, the signal envelope tilt is converted into a voice coil offset using a pre-trained polynomial function. The general form of the polynomial transformation is y = k0 + k1x + k2x 2 +…+k N x N

[0090] The symbols are annotated as follows:

[0091] y: Voice coil offset (unit: millimeters).

[0092] x: represents the skewness of the signal envelope, and is a scalar value.

[0093] k0,k1,k2,…,k N : Polynomial coefficients determined through training.

[0094] N: The order of the polynomial, which is a positive integer.

[0095] For example, using the model y=k0+k1*x+k2*x^2+…+k N *x^N.

[0096] S700 (optional): Quality assessment.

[0097] After obtaining the voice coil offset, it can be compared with a preset tolerance range (e.g., ±0.5 mm). This tolerance range is preset based on the product design requirements and performance standards of this speaker model and is typically determined through tests that correlate the voice coil offset with audio performance indicators (such as total harmonic distortion). Finally, a pass / fail result is output.

[0098] Those skilled in the art will understand that the above embodiments are merely one implementation. Based on the technical principles, the necessary steps of this method include acquiring the measurement signal S100, calculating the signal envelope S200, calculating the envelope bias S300, calculating the envelope tilt S500, and transforming it into a voice coil offset S600. Normalizing the envelope bias S400 and quality judgment S700 are optional steps used to improve robustness and achieve automated quality inspection. When the measurement environment is stable, step S400 can be omitted, and subsequent calculations can be performed directly based on the original envelope bias.

[0099] Through the above process, this invention achieves rapid, automated, and high-precision online detection of loudspeaker voice coil offset. This technology is suitable for production quality control and production process verification. It can be applied as a software module in loudspeaker product measurement solutions. It accesses signals acquired by the measurement system and processes these signals to estimate the loudspeaker voice coil offset. Configurable limits can be set to determine whether the current loudspeaker meets the acceptable range for voice coil offset. When the loudspeaker voice coil offset value is within the defined acceptable range, the system indicates that the loudspeaker meets the established quality standards; otherwise, it indicates that it does not meet the quality standards.

[0100] This invention also proposes a loudspeaker manufacturing quality control system, such as... Figure 9 As shown, the system includes a measurement system and a processor. The measurement system is used to acquire the measurement signal of the loudspeaker under test, and the processor operates using the method of estimating the loudspeaker voice coil offset based on the signal envelope as described in any of the embodiments above. Since this system adopts all the technical solutions of all the embodiments of the above methods, it has at least all the beneficial effects brought about by the technical solutions of the above embodiments, which will not be repeated here.

[0101] The present invention also proposes a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform the method for estimating loudspeaker voice coil offset based on signal envelope as described in any of the above embodiments. Since this computer-readable storage medium employs all the technical solutions of all the embodiments described above, it possesses at least all the beneficial effects brought about by the technical solutions of the above embodiments, which will not be elaborated upon here.

[0102] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method of estimating speaker voice coil excursion based on signal envelope, the method comprising: The method comprises the following steps: obtaining a measurement signal of a speaker to be tested; calculating an upper envelope and a lower envelope of the measurement signal; calculating a signal envelope bias based on the upper envelope and the lower envelope; calculating a signal envelope tilt based on the signal envelope bias; converting the signal envelope tilt into a voice coil excursion of the speaker to be tested through a polynomial transformation.

2. The method of claim 1, wherein, After calculating the signal envelope bias, before calculating the signal envelope tilt, the method further comprises the steps of: obtaining a reference signal envelope bias of a reference speaker; subtracting the signal envelope bias of the speaker to be tested from the reference signal envelope bias to obtain a normalized signal envelope bias; the step of calculating the signal envelope tilt is based on the normalized signal envelope bias.

3. The method of claim 1 or 2, wherein, The step of calculating the signal envelope tilt comprises the steps of: dividing the signal envelope bias or the normalized signal envelope bias into a low-frequency segment and a high-frequency segment according to a resonance frequency of the speaker; calculating the algebraic sum of the envelope bias in the low-frequency segment and the high-frequency segment respectively, and selecting the maximum value or the minimum value of the envelope bias in the low-frequency segment and the high-frequency segment respectively according to the positive and negative of the algebraic sum; calculating the signal envelope tilt by using eigenvalues, and the signal envelope tilt is the algebraic difference between a low-frequency eigenvalue and a high-frequency eigenvalue, and the calculation formula of the signal envelope tilt is: x = feature low -feature high where x is the signal envelope tilt and is a scalar value;feature low is the low frequency feature value;feature high is the high frequency feature value.

4. The method of claim 1, wherein, The polynomial transformation is in the form of: y = k0+ k1x + k2x 2 +... + k N x N where y is the voice coil excursion; x is the signal envelope tilt and is a scalar value; k0, k1, …, k N are polynomial coefficients; N is the polynomial order.

5. The method of claim 4, wherein, The polynomial coefficients are determined through a training process, and the training process comprises the steps of: obtaining measurement signals of a plurality of speaker samples with known voice coil excursions; calculating corresponding signal envelope tilts according to the measurement signals; using a regression algorithm to fit a polynomial relationship between the known voice coil excursions and the signal envelope tilts, thereby determining the polynomial coefficients.

6. The method of claim 5, wherein, The speaker samples with known voice coil excursions are obtained by at least one of the following methods: performing computer physics simulation on speakers with different voice coil excursions; manufacturing physical speaker samples with different voice coil excursions; collecting speaker data in production trial operation; applying different sizes of direct current signals to standard speaker samples to simulate different voice coil excursions.

7. The method of claim 1, wherein, The measurement signal is a time-domain signal, and the types thereof include pressure signals, acceleration signals or displacement signals.

8. The method of claim 1, wherein, After obtaining the voice coil excursion, the method further comprises the steps of: comparing the voice coil excursion with a preset tolerance range; outputting a judgment conclusion that the speaker to be tested is qualified or unqualified according to the comparison result.

9. A loudspeaker production quality control system comprising a measurement system and a processor, characterised in that, The measurement system is used to obtain the measurement signal of the speaker to be tested, and the processor works by using the method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a stored computer program, wherein the computer readable storage medium controls the device or apparatus where the computer readable storage medium is located to execute the method according to any one of claims 1 to 8 when the computer program runs.

Images