ADC metastable state elimination method based on field programmable gate array

Abstract

The invention discloses an ADC (Analog to Digital Converter) metastable state elimination method based on a field programmable gate array. The ADC metastable state elimination method comprises the following steps: acquiring multi-dimensional feature data of sampled ADC data, which reflects metastable state features in multiple dimensions of time domain, frequency domain and consistency; based on the multi-dimensional feature data, determining an ADC sampling data state; and if the ADC sampling data is in a metastable state, adjusting a channel associated clock of the ADC by adopting a delay chain of the programmable gate array to obtain stable ADC sampling data. According to the method, the multi-dimensional feature data reflecting the metastable state features of the sampled ADC data in multiple dimensions such as the time domain, the frequency domain and the consistency are acquired, and the ADC sampling data state is determined based on the multi-dimensional feature data, so that the metastable state is accurately identified, and after the metastable state is identified, the metastable state is accurately identified. The delay chain of the programmable gate array is adopted to timely and effectively adjust the channel associated clock of the ADC, and the metastable state can be quickly eliminated.

Description

Technical Field

[0001] The present invention relates to a method for eliminating metastability of an ADC based on a field programmable gate array, belonging to the technical field of digital signal processing. Background Art

[0002] Metastability means that when the digital signal output by an analog-to-digital converter (ADC) jumps at the edge of the clock sampling window of a field programmable gate array (FPGA), the FPGA may not be able to deterministically recognize the signal as logic "0" or logic "1", resulting in uncertain and random errors in the captured data; such errors usually manifest as sudden jumps in data, abnormal value distributions, or unstable fluctuations in the steady-state level, seriously affecting the measurement accuracy and reliability of the system.

[0003] The prior art can meet the requirements of metastability detection and phase self-optimization to a certain extent, but there are still problems such as insufficient detection accuracy, low sensitivity, and increased bandwidth occupancy; especially in occasions where bandwidth resources are scarce and high data accuracy is required, the existing detection and optimization methods are difficult to provide sufficient reliability. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method for eliminating metastability of an ADC based on a field programmable gate array, which can accurately identify the timing of metastability occurrence in the ADC and eliminate metastability in a timely manner, facilitating the improvement of the measurement progress of ADC sampling data, and solving the problems of the current inability to detect metastability phenomena occurring during ADC sampling in a timely manner and the inability to effectively adjust metastability.

[0005] To achieve the above object / To solve the above technical problems, the present invention is implemented by the following technical solutions: A method for eliminating metastability of an ADC based on a field programmable gate array, comprising: Obtaining multi-dimensional feature data reflecting metastability characteristics of the sampled ADC data in multiple dimensions of time domain, frequency domain, and consistency; Determining the state of the ADC sampling data based on the multi-dimensional feature data; If the state of the ADC sampling data is metastable, the delay chain of the programmable gate array is used to adjust the clock associated with the ADC to obtain stable ADC sampling data.

[0006] Further, the obtaining of the multi-dimensional feature data reflecting metastability characteristics of the sampled ADC data in multiple dimensions of time domain, frequency domain, and consistency includes: Performing feature extraction on the sampled ADC data to obtain the jump rate of time domain features, the high-frequency energy ratio of frequency domain features, and the coefficient of variation of consistency features; Perform feature fusion processing on the jump rate of time-domain features, the high-frequency energy proportion of frequency-domain features, and the coefficient of variation of consistency features to obtain multi-dimensional feature data.

[0007] Furthermore, calculate the jump rate of time-domain features through the following formula: ; ; ; where: N represents the total number of sampling points, n represents the sampling point number in the time domain; represents the amplitude difference between two adjacent sampling points, represents the signal amplitude of the n th sampling point in the time domain, represents the signal amplitude of the n +1th sampling point in the time domain; represents the jump rate of time-domain features; represents the preset jump threshold, represents the average value of the differences between adjacent sampling points during normal sampling, represents the standard deviation of the differences between two adjacent sampling points during normal sampling; represents the jump threshold coefficient.

[0008] Furthermore, calculate the high-frequency energy proportion of frequency-domain features through the following formula: ; ; where: represents the high-frequency energy proportion of frequency-domain features; N represents the total number of sampling points, n represents the sampling point number in the time domain, i represents the frequency component number in the frequency domain; represents the fast Fourier transform, represents the signal amplitude of the n th sampling point in the time domain; represents the complex number of the i th frequency component in the frequency domain; k represents the high-frequency threshold.

[0009] Further, the coefficient of variation of the consistency feature is calculated by the following formula: ; Where: represents the coefficient of variation of the consistency feature; represents the standard deviation of the sampling sequence, represents the mean of the sampling sequence, represents the n th sampling sequence.

[0010] Further, the jump rate of the time-domain feature, the high-frequency energy ratio of the frequency-domain feature, and the coefficient of variation of the consistency feature are subjected to feature fusion processing to obtain multi-dimensional feature data, including: The multi-dimensional feature data is calculated by the following formula: ; Where: represents the multi-dimensional feature data; represents the normalized jump rate, represents the weight parameter of the normalized jump rate; represents the normalized high-frequency energy ratio, represents the weight parameter of the normalized high-frequency energy ratio; represents the normalized coefficient of variation, represents the weight parameter of the normalized coefficient of variation; , .

[0011] Further, determining the ADC sampling data state based on the multi-dimensional feature data includes: If the multi-dimensional feature data is greater than the preset judgment threshold, the ADC sampling data is metastable, otherwise the ADC data sampling is normal.

[0012] Further, the preset judgment threshold is calculated by the following formula: ; Where: represents the preset judgment threshold; represents the mean of the multi-dimensional feature data normalized by multiple normal samplings, represents the standard deviation of the multi-dimensional feature data normalized by multiple normal samplings; represents the basic safety margin coefficient, with a value range of 2.5 to 3.5; represents the adaptive factor, with a value range of 0.2 to 0.4; represents the maximum value in the multi-dimensional feature data obtained by normalizing multiple normal samplings; represents the minimum value in the multi-dimensional feature data obtained by normalizing multiple normal samplings.

[0013] Furthermore, if the ADC sampling data is metastable, the delay chain of the programmable gate array is used to adjust the clock of the ADC on the same path to obtain stable ADC sampling data, including: The phase of the clock of the ADC on the same path is adjusted by using the IDLEAY2 delay chain of the programmable gate array. Among them, all tap coefficients are traversed within the preset tap coefficient range of the clock delay to obtain a set of tap coefficient sequences corresponding to the non-metastable ADC sampling data; The median of the tap coefficient sequence is taken as the tap coefficient of the final sampling clock delay to obtain stable ADC sampling data.

[0014] Furthermore, the use of the delay chain of the programmable gate array to adjust the clock of the ADC on the same path includes: The process of phase adjustment is represented by the following formula: ; ; Where: represents the adjusted clock delay; represents the tap coefficient of the clock delay, with a value range of 0 to 31; represents the delay adjustment step size, represents the reference clock frequency, represents the device parameters of the programmable gate array.

[0015] Compared with the prior art, the beneficial effects achieved by the present invention are: 1. The present invention realizes the accurate identification of metastability by obtaining multi-dimensional feature data that reflects the metastable characteristics of the sampled ADC data in multiple dimensions of time domain, frequency domain, and consistency, and determining the state of the ADC sampling data based on the multi-dimensional feature data. After identifying the metastability, the delay chain of the programmable gate array is used to timely and effectively adjust the clock of the ADC on the same path, which can quickly eliminate the metastability, solve the problem that the current metastable phenomenon in the ADC sampling process cannot be detected in time, and the metastability cannot be effectively adjusted.

[0016] 2. The present invention uses the IDLEAY2 delay chain of the programmable gate array to adjust the phase of the clock associated with the ADC. By traversing all tap coefficients within the preset clock delay tap coefficient range, a set of tap coefficient sequences corresponding to the metastability-free ADC sampling data is obtained. The median of the tap coefficient sequences is taken as the tap coefficient for the final sampling clock delay, which can quickly eliminate metastability and obtain stable ADC sampling data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flowchart of the method for eliminating ADC metastability based on a field programmable gate array provided by an embodiment of the present invention; Figure 2 is a flowchart of obtaining multi-dimensional feature data provided by an embodiment of the present invention; Figure 3 is a schematic diagram of obtaining multi-dimensional feature data provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be used to limit the protection scope of the present invention. Embodiment

[0019] As Figure 1 shown, the method for eliminating ADC metastability based on a field programmable gate array includes: Obtaining multi-dimensional feature data reflecting metastability characteristics of the sampled ADC data in multiple dimensions of time domain, frequency domain, and consistency. Specifically: As Figure 2 shown, performing feature extraction on the sampled ADC data to obtain the jump rate of the time domain feature, the high-frequency energy ratio of the frequency domain feature, and the coefficient of variation of the consistency feature; Calculating the jump rate of the time domain feature through the following formula: ; ; ; Where: N represents the total number of sampling points, n represents the sampling point number in the time domain; represents the amplitude difference between two adjacent sampling points, represents the signal amplitude of the n th sampling point in the time domain, represents the n +1th sampling point in the time domain; Represents the jump rate of time-domain features; Represents a preset jump threshold, Represents the average value of the differences between adjacent sampling points during normal sampling, Represents the standard deviation of the differences between adjacent two sampling points during normal sampling; Represents the jump threshold coefficient, The value is 3; Calculate the high-frequency energy ratio of frequency-domain features through the following formula: ; ; Represents the high-frequency energy ratio of frequency-domain features; N Represents the total number of sampling points, n Represents the sampling point sequence number in the time domain, i Represents the frequency component sequence number in the frequency domain; Represents the fast Fourier transform, Represents the signal amplitude of the n th sampling point in the time domain; Represents the complex number of the i th frequency component in the frequency domain; k Represents the high-frequency threshold; Calculate the coefficient of variation of the consistency feature through the following formula: ; Represents the coefficient of variation of the consistency feature; Represents the standard deviation of the sampling sequence, Represents the mean value of the sampling sequence, Represents the n th sampling sequence.

[0020] Perform feature fusion processing on the jump rate of time-domain features, the high-frequency energy ratio of frequency-domain features, and the coefficient of variation of consistency features to obtain multi-dimensional feature data; Calculate the multi-dimensional feature data using the following formula: ; Represents the multi-dimensional feature data; Represents the normalized jump rate, Represents the weight parameter of the normalized jump rate; Represents the normalized high-frequency energy ratio, A weight parameter representing the proportion of normalized high-frequency energy; Represents the normalized coefficient of variation, A weight parameter representing the normalized coefficient of variation; The coefficient of variation can complement the jump rate and the proportion of high-frequency energy. The three weight parameters should satisfy the following relationship: , .

[0021] Based on multi-dimensional feature data, determine the state of ADC sampling data. Specifically: If the multi-dimensional feature data is greater than the preset judgment threshold, the ADC sampling data is metastable; otherwise, the ADC data sampling is normal. Among them, the preset judgment threshold is calculated by the following formula: ; Represents the preset judgment threshold; Represents the mean value of the multi-dimensional feature data normalized by multiple normal samplings, Represents the standard deviation of the multi-dimensional feature data normalized by multiple normal samplings; Represents the basic safety margin coefficient, with a value range of 2.5 to 3.5; Represents the adaptive factor, with a value range of 0.2 to 0.4; Represents the maximum value in the multi-dimensional feature data normalized by multiple normal samplings; Represents the minimum value in the multi-dimensional feature data normalized by multiple normal samplings.

[0022] If the ADC sampling data is metastable, the delay chain of the programmable gate array is used to adjust the clock associated with the ADC.

[0023] Embodiment 2 As Figure 1 shown, the ADC metastability elimination method based on a field programmable gate array includes: Obtain multi-dimensional feature data reflecting the metastability characteristics of the sampled ADC data in multiple dimensions of time domain, frequency domain, and consistency. Specifically: As Figure 2 and Figure 3 shown, perform feature extraction on the sampled ADC data to obtain the jump rate of the time domain feature, the proportion of high-frequency energy of the frequency domain feature, and the coefficient of variation of the consistency feature; Calculate the jump rate of the time domain feature by the following formula: ; ; ; Wherein: N represents the total number of sampling points, n represents the sampling point sequence number in the time domain; represents the amplitude difference between two adjacent sampling points, represents the n th sampling point signal amplitude in the time domain, represents the n +1th sampling point signal amplitude in the time domain; represents the jump rate of the time domain feature; represents a preset jump threshold, represents the average value of the differences between adjacent sampling points during normal sampling, represents the standard deviation of the differences between two adjacent sampling points during normal sampling; represents the jump threshold coefficient, The value is 3; The high-frequency energy ratio of the frequency domain feature is calculated by the following formula: ; ; represents the high-frequency energy ratio of the frequency domain feature; N represents the total number of sampling points, n represents the sampling point sequence number in the time domain, i represents the frequency component sequence number in the frequency domain; represents the fast Fourier transform, represents the n th sampling point signal amplitude in the time domain; represents the i th complex number of the frequency component in the frequency domain; k represents the high-frequency threshold; The coefficient of variation of the consistency feature is calculated by the following formula: ; represents the coefficient of variation of the consistency feature; represents the standard deviation of the sampling sequence, represents the mean value of the sampling sequence, represents the n th sampling sequence.

[0024] Perform feature fusion processing on the jump rate of time-domain features, the high-frequency energy ratio of frequency-domain features, and the coefficient of variation of consistency features to obtain multi-dimensional feature data; The multi-dimensional feature data is calculated using the following formula: ; represents the multi-dimensional feature data; represents the normalized jump rate, represents the weight parameter of the normalized jump rate; represents the normalized high-frequency energy ratio, represents the weight parameter of the normalized high-frequency energy ratio; represents the normalized coefficient of variation, represents the weight parameter of the normalized coefficient of variation; The coefficient of variation can complement the jump rate and the high-frequency energy ratio, and the three weight parameters should satisfy the following relationship: , ; , represents the reference mean of the jump rate calculated based on a large number of normal sampling data samples under a specified sampling scenario, represents the reference standard deviation of the jump rate calculated based on a large number of normal sampling data samples under a specified sampling scenario; F represents the calibration factor; , represents the reference mean of the high-frequency energy calculated based on a large number of normal sampling data samples under a specified sampling scenario, represents the reference standard deviation of the high-frequency energy calculated based on a large number of normal sampling data samples under a specified sampling scenario; , represents the reference mean of the coefficient of variation calculated based on a large number of normal sampling data samples under a specified sampling scenario, represents the reference standard deviation of the coefficient of variation calculated based on a large number of normal sampling data samples under a specified sampling scenario.

[0025] Based on the multi-dimensional feature data, determine the state of ADC sampling data. Specifically: If the multi-dimensional feature data is greater than the preset judgment threshold, the ADC sampling data is metastable; otherwise, the ADC data sampling is normal. Among them, the preset judgment threshold is calculated using the following formula: ; represents the preset judgment threshold; Represents the mean of the multi-dimensional feature data after multiple normal samplings and normalization. Represents the standard deviation of the multi-dimensional feature data after multiple normal samplings and normalization. Represents the basic safety margin coefficient, with a value range of 2.5 to 3.5. Represents the adaptive factor, with a value range of 0.2 to 0.4. Represents the maximum value in the multi-dimensional feature data after multiple normal samplings and normalization. Represents the minimum value in the multi-dimensional feature data after multiple normal samplings and normalization.

[0026] If the ADC sampling data is metastable, the delay chain of the programmable gate array is used to adjust the clock of the ADC's companion clock. Specifically: The IDLEAY2 delay chain of the programmable gate array is used to adjust the phase of the ADC's companion clock. Among them, all tap coefficients are traversed within the preset clock delay tap coefficient range to obtain a set of tap coefficient sequences corresponding to non-metastable ADC sampling data. The median of the tap coefficient sequence is taken as the tap coefficient of the final sampling clock delay to obtain stable ADC sampling data. Among them, the following formula is used to represent the phase adjustment process: ; ; Represents the adjusted clock delay. Represents the tap coefficient of the clock delay, with a value range of 0 to 31. Represents the delay adjustment step size. Represents the reference clock frequency. Represents the device parameter of the programmable gate array, with a value of 64 in the high-resolution mode and 32 in the low-resolution mode. In this embodiment, a high-resolution field programmable gate array (FPGA) device is used, with 200 Mhz as the reference clock frequency. Then, the delay adjustment step size each time satisfies that the rising edge of the companion clock sampling does not fall within the setup and hold time of the ADC sampling data, thus avoiding metastability.

[0027] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0028] The present application is described with reference to the flowcharts of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process in the flowchart can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes or blocks Figure 1 or multiple blocks.

[0029] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes.

[0030] These computer program instructions can also be loaded onto a computer or other programmable data processing devices, so that a series of operation steps are executed on the computer or other programmable devices to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable devices provide steps for implementing the functions specified in Figure 1 one process or multiple processes.

[0031] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit and scope protected by the present invention and the claims. These all belong to the protection scope of the present invention.

Claims

1. A method for eliminating metastability of an ADC based on a field programmable gate array, characterized in that Including: Obtaining multi-dimensional feature data that reflects metastable characteristics of sampled ADC data in multiple dimensions of time domain, frequency domain, and consistency; Determining the state of ADC sampling data based on the multi-dimensional feature data; If the state of the ADC sampling data is metastable, adjusting the clock associated with the ADC using the delay chain of the programmable gate array to obtain stable ADC sampling data.

2. The method for eliminating ADC metastability based on a field programmable gate array according to claim 1, wherein The obtaining of multi-dimensional feature data that reflects metastable characteristics of sampled ADC data in multiple dimensions of time domain, frequency domain, and consistency includes: Performing feature extraction on the sampled ADC data to obtain the jump rate of the time-domain feature, the high-frequency energy ratio of the frequency-domain feature, and the coefficient of variation of the consistency feature; Performing feature fusion processing on the jump rate of the time-domain feature, the high-frequency energy ratio of the frequency-domain feature, and the coefficient of variation of the consistency feature to obtain multi-dimensional feature data.

3. The ADC metastability elimination method based on a field programmable gate array according to claim 2, wherein Calculating the jump rate of the time-domain feature through the following formula: ； ； ； Where: N represents the total number of sampling points, n represents the sampling point sequence number in the time domain; represents the amplitude difference between two adjacent sampling points, represents the signal amplitude at the n -th sampling point in the time domain, represents the signal amplitude at the n +1-th sampling point in the time domain; Indicates the jump rate of time-domain features; represents a preset jump threshold, represents the average value of the differences between adjacent sampling points during normal sampling, represents the standard deviation of the differences between two adjacent sampling points during normal sampling; Indicates the jump threshold coefficient.

4. The ADC metastability elimination method based on field programmable gate array according to claim 2, wherein Calculating the high-frequency energy ratio of the frequency-domain feature through the following formula: ； ； Where: Indicates the proportion of high-frequency energy in the frequency-domain feature; N represents the total number of sampling points, n represents the sampling point sequence number in the time domain, i represents the frequency component sequence number in the frequency domain; represents the Fast Fourier Transform, represents the signal amplitude of the n th sampling point in the time domain; Denote the complex number of the i th frequency component in the frequency domain; k Indicates the high-frequency threshold.

5. The ADC metastability elimination method based on a field programmable gate array according to claim 2, wherein Calculating the coefficient of variation of the consistency feature through the following formula: ； Where: Coefficient of variation representing the consistency feature; Standard deviation representing the sampling sequence, Mean value representing the sampling sequence, Indicating the n th sampling sequence.

6. The ADC metastability elimination method based on field programmable gate array according to claim 2, characterized in that, The performing of feature fusion processing on the jump rate of the time-domain feature, the high-frequency energy ratio of the frequency-domain feature, and the coefficient of variation of the consistency feature to obtain multi-dimensional feature data includes: calculating the multi-dimensional feature data using the following formula: ； Where: Represents multi-dimensional feature data; represents the normalized jump rate, and represents the weight parameter of the normalized jump rate; represents the proportion of normalized high-frequency energy, is the weight parameter representing the proportion of normalized high-frequency energy; represents the normalized coefficient of variation, represents the weight parameter of the normalized coefficient of variation; ， 。 7. The method for eliminating metastability of an ADC based on a field programmable gate array according to claim 1, wherein The determining of the state of ADC sampling data based on the multi-dimensional feature data includes: If the multi-dimensional feature data is greater than the preset judgment threshold, the ADC sampling data is metastable; otherwise, the ADC data sampling is normal.

8. The method for eliminating metastability of ADC based on field programmable gate array according to claim 7, characterized in that, Calculating the preset judgment threshold through the following formula: ； Where: denotes a preset judgment threshold value; represents the mean of the multi-dimensional feature data after multiple normal samplings and normalizations, represents the standard deviation of the multi-dimensional feature data after multiple normal samplings and normalizations; Indicates the basic safety margin coefficient, with a value range of 2.5 to 3.5; Represents the adaptive factor, with a value range of 0.2 to 0.4; Represents the maximum value in the multi-dimensional feature data after multiple normal samplings and normalizations; Represents the minimum value in the multi-dimensional feature data after multiple normal samplings and normalizations.

9. The method for eliminating metastability of ADC based on field programmable gate array according to claim 7, wherein The if the ADC sampling data is metastable, adjusting the clock associated with the ADC using the delay chain of the programmable gate array to obtain stable ADC sampling data includes: Performing phase adjustment on the clock associated with the ADC using the IDLEAY2 delay chain of the programmable gate array, where all tap coefficients are traversed within the preset tap coefficient range of the clock delay to obtain a set of tap coefficient sequences corresponding to the ADC sampling data without metastability; Taking the median of the tap coefficient sequences as the tap coefficient of the final sampling clock delay to obtain stable ADC sampling data.

10. The method for eliminating ADC metastability based on field programmable gate array according to claim 9, characterized in that, The adjusting of the clock associated with the ADC using the delay chain of the programmable gate array includes: representing the process of phase adjustment using the following formula: ； ； Where: Indicates the adjusted clock delay; Indicates the tap coefficient for clock delay, with values ranging from 0 to 31; represents the delay adjustment step size, represents the reference clock frequency, represents the device parameters of the programmable gate array.

Images