An ADC digital background calibration method based on intelligent optimization algorithm

Abstract

The present invention discloses a method for ADC digital background calibration based on an intelligent optimization algorithm, comprising: after receiving M groups of output data sent by an analog-to-digital converter, constructing M groups of interpolation output codes using a Lagrange interpolation algorithm; constructing a target optimization function based on the M groups of output data and the M groups of interpolation output codes, and performing an optimization search using a particle swarm optimization algorithm to obtain M groups of weights so that the mean square error between each group of output data and its corresponding interpolation output code is minimized; and the average value of the M groups of weights is output as the optimal weight for this optimization search and as the central value for the next optimization search. By combining the interpolation and particle swarm optimization algorithms as a framework to optimize the search for the target optimization function, with the goal of minimizing the mean square error between the M groups of output data and their corresponding interpolation output codes, the optimal weight is obtained, thereby improving the accuracy of the successive approximation ADC and simply and efficiently implementing the digital background calibration of the ADC.

Description

Technical Field

[0001] The present invention belongs to the technical field of integrated circuits, and in particular relates to an ADC digital background calibration method based on an intelligent optimization algorithm. Background Art

[0002] With the development of integrated circuit technology, the design difficulty of analog integrated circuits continues to increase. The successive approximation analog-to-digital converter is a relatively process-friendly analog-to-digital converter architecture, but capacitor mismatch will significantly affect the accuracy and linearity of this analog-to-digital converter.

[0003] Existing digital background calibration methods, such as those based on a benchmark ADC (Analog-to-Digital Converter), split ADC calibration, and PN pseudo-random code injection, can calibrate ADC errors. However, these methods fail to incorporate empirical knowledge of ADCs into the calibration algorithms. This results in the potential for the algorithms to not converge or to converge to incorrect solutions. To avoid this, complex algorithm convergence mechanisms and analog circuit modules are often required to ensure algorithm reliability, significantly increasing the difficulty of calibration. Summary of the Invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides an ADC digital background calibration method based on an intelligent optimization algorithm. The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0005] The present invention provides an ADC digital background calibration method based on an intelligent optimization algorithm, comprising:

[0006] After receiving M groups of output data sent by the analog-to-digital converter, construct M groups of interpolation output codes using the Lagrange interpolation algorithm;

[0007] A target optimization function is constructed based on the M groups of output data and the M groups of interpolation output codes, and a particle swarm optimization algorithm is used to optimize and search the target optimization function to obtain M groups of weights so that the mean square error between each group of output data and its corresponding interpolation output code is minimized;

[0008] The average value of the M groups of weights is output as the optimal weight for this optimization search.

[0009] In one embodiment of the present invention, after receiving M groups of output data sent by the analog-to-digital converter, the step of constructing M groups of interpolation output codes using a Lagrange interpolation algorithm includes:

[0010] receiving M groups of output data sent by the analog-to-digital converter;

[0011] Determine the order of the Lagrange interpolation algorithm used for fitting interpolation: Wherein, N represents the digital output bit width of the analog-to-digital converter is N bits, Indicates rounding down;

[0012] For each set of output data, an interpolation output code is constructed according to the order of the fitting interpolation using a Lagrange interpolation algorithm to obtain M sets of interpolation output codes corresponding to the M sets of output data.

[0013] In one embodiment of the present invention, before the step of constructing the interpolation output code according to the order of the fitting interpolation using the Lagrange interpolation algorithm for each set of the output data, the method further includes:

[0014] Capacitance mismatch in the analog-to-digital converter is estimated, and an initial search range of the particle swarm optimization algorithm is set when optimizing each set of output data.

[0015] In one embodiment of the present invention, after the step of outputting the average value of the M groups of weights as the optimal weight for this optimization search, the method further includes:

[0016] Returning to the step of receiving M groups of output data sent by the analog-to-digital converter, and determining an initial search range of the particle swarm optimization algorithm for the next optimization search of each group of output data and the interpolation output code according to the average value;

[0017] Wherein, the initial search range of the particle swarm optimization algorithm during the next optimization search takes the average value as the center value.

[0018] In one embodiment of the present invention, the objective optimization function is:

[0019]

[0020] Among them, W represents the weight to be optimized, D j represents the actual output of the j-th iteration of each set of output data, represents the output of the interpolation, k represents the preset number of iterations for each set of output data, D j and The dimensions of are 1×N, and the dimensions of W are N×1.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] The present invention provides an ADC digital background calibration method based on an intelligent optimization algorithm. By using an interpolation particle swarm optimization algorithm as a framework to optimize and search for a target optimization function, with the goal of minimizing the mean square error between each set of output data and its corresponding interpolated output code, the optimal weight of the optimization search is obtained, thereby improving the accuracy of a successive approximation ADC and simply and efficiently achieving calibration of the ADC digital background.

[0023] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a flow chart of an ADC calibration method based on an intelligent optimization algorithm provided by an embodiment of the present invention;

[0025] Figure 2 1 is a schematic diagram of an ADC calibration method based on an intelligent optimization algorithm provided by an embodiment of the present invention;

[0026] Figure 3 1 is a schematic diagram of the dynamic characteristics of the SAR ADC before calibration provided by an embodiment of the present invention;

[0027] Figure 4 FIG. 4 is a schematic diagram of the dynamic characteristics of the SAR ADC after calibration provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The present invention will be further described in detail below with reference to specific examples, but the embodiments of the present invention are not limited thereto.

[0029] Figure 1 This is a flow chart of an ADC calibration method based on an intelligent optimization algorithm provided by an embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides an ADC digital background calibration method based on an intelligent optimization algorithm, comprising:

[0030] S1, after receiving M groups of output data sent by the analog-to-digital converter, construct M groups of interpolation output codes using the Lagrange interpolation algorithm;

[0031] S2. Construct a target optimization function based on the M groups of output data and the M groups of interpolation output codes, and use the particle swarm optimization algorithm to optimize the target optimization function to obtain M groups of weights so that the mean square error between each group of output data and its corresponding interpolation output code is minimized;

[0032] S3. Output the average value of the M groups of weights as the optimal weight for this optimization search.

[0033] Optionally, in the above step S1, after receiving M groups of output data sent by the analog-to-digital converter, the step of constructing M groups of interpolation output codes using a Lagrange interpolation algorithm includes:

[0034] S101, receiving M groups of output data sent by an analog-to-digital converter;

[0035] S102, determining the order of fitting interpolation using the Lagrange interpolation algorithm: Where N represents the digital output bit width of the analog-to-digital converter, which is N bits. Indicates rounding down;

[0036] S103 , for each set of output data, construct an interpolation output code according to the order of the fitted interpolation using the Lagrange interpolation algorithm to obtain M sets of interpolation output codes corresponding to the M sets of output data.

[0037] Specifically, in order to make the interpolation error within 1LSB (Least Significant Bit), the formula can be used before using the Lagrange algorithm to interpolate Determine the order of the fitting interpolation, where N represents the digital output bit width of the analog-to-digital converter, which is N bits. Optionally, this embodiment selectively uses an analog-to-digital converter with a digital output bit width of 12-16 bits. Taking an analog-to-digital converter with a digital output bit width of 14 bits as an example, the order of the fitting interpolation is 5.

[0038] Furthermore, in the above step S103, for an N-bit analog-to-digital converter, Lagrange interpolation is performed on each set of output data according to the following formula:

[0039]

[0040] Where D j represents the output data of the jth iteration in the optimization search process of each group of output data, h j Represents the preset interpolation coefficients for the jth iteration in the optimization search process for each set of output data.

[0041] Optionally, before the step of constructing the interpolation output code according to the order of the fitting interpolation using the Lagrange interpolation algorithm for each set of output data, the method further includes:

[0042] The capacitor mismatch in the analog-to-digital converter is estimated and the initial search range of the particle swarm optimization algorithm is set when optimizing each set of output data.

[0043] It should be understood that estimating the capacitor mismatch in the ADC can be achieved through simulation, searching process library files, and other means. Since the optimization search for each set of data and interpolated output code requires multiple iterations, the estimated capacitor mismatch can be used as prior knowledge to set the initial search range of the particle swarm optimization algorithm during the first iteration of each set of data. This design eliminates some erroneous solutions that may lead to convergence, ensures the accuracy of the particle swarm optimization algorithm, and makes the obtained solution closer to the actual weight.

[0044] Please continue to see Figure 1 After the above step of outputting the average value of the M groups of weights as the optimal weight for this optimization search, the following steps are further included:

[0045] S4, returning to the step of receiving M groups of output data sent by the analog-to-digital converter, and determining the initial search range of the particle swarm optimization algorithm for the next optimization search of each group of output data and the interpolation output code according to the average value;

[0046] Among them, the initial search range of the particle swarm optimization algorithm in the next optimization search takes the average value as the center value.

[0047] Optionally, the objective optimization function is:

[0048]

[0049] Among them, W represents the weight to be optimized, D j represents the actual output of the j-th iteration of each set of output data, represents the interpolation output code of the jth iteration during the optimization search process of each group of output data, k represents the preset number of iterations during the optimization search, D j and The dimensions of are 1×N, and the dimensions of W are N×1.

[0050] The following takes a SAR ADC with a calibration accuracy of 14 bits as an example to further illustrate the ADC calibration method based on the intelligent optimization algorithm. Figure 2 As shown in the figure, for a 14-bit SAR ADC, when the order of Lagrange interpolation is 5, the error can be less than 1LSB, so the interpolation expression is:

[0051]

[0052] Where h is the interpolation coefficient, h j Take 0.05, -0.3, 0.75, 0.75, -0.3 and 0.05 in turn.

[0053] Next, the interpolated output code and output data are used to establish an optimization objective function to provide an error convergence criterion. The specific expression of the optimization objective function is:

[0054]

[0055] Furthermore, the particle swarm optimization algorithm is used to solve the actual weight of the SAR ADC. The particle swarm optimization algorithm has the advantages of simple structure and easy implementation. Specifically, the weight value of the SAR ADC is used as the search position of the particle swarm optimization algorithm. The capacitor mismatch is estimated by simulation and searching the process library file. This is used as prior knowledge to set the initial search range of the particle swarm optimization algorithm during the optimization search. In this embodiment, if the ratio of the mean square value σ of the capacitor process error used to the capacitance value C is 3%, then the center value of the initial search range during the optimization search is a set of ideal binary numbers: W' = [2^(-1)2^(-2)2^(-3)2^(-4)2^(-5)2^(-6)2^(-7)2^(-8)2^(-9)2^(-10)2^(-11)2^(-12)2^(-13)2^(-14)].

[0056] For example, for the upper 10 bits of the digital output bit width, the search range is set to W'±20LSB, the minimum value of the lower four bits is set to 0, and the maximum value is set to W'±5LSB.

[0057] In this embodiment, the real weight of the SAR ADC is calculated using a particle swarm optimization algorithm with linearly decreasing inertia weights. The particle swarm optimization algorithm with linearly decreasing inertia weights can be described as:

[0058] v i (t+1)=wv i (t)+c1r1(p i -x i (t))+c2r2(p g -x i (t))

[0059] x i (t+1)=x i (t)+v i (t+1)

[0060]

[0061] Among them, x i and v i Represent the position vector and velocity vector of the i-th particle, w represents the inertia weight, p i represents the local optimal position, p g Indicates the global optimal position, c1 and c2 are constants, r1 and r2 are random numbers between [0,1], iter indicates the current number of iterations, iter max Indicates the maximum number of iterations, wmax and w min Represent the maximum and minimum values ​​of the inertia weight respectively.

[0062] The process of calculating the true weight of SAR ADC using the particle swarm optimization algorithm with linearly decreasing inertia weight is as follows:

[0063] (1) Initialize the parameters of the particle swarm optimization algorithm, as shown in Table 1:

[0064] Table 1

[0065]

[0066]

[0067] (2) After receiving M sets of output data, the error between the output interpolation code and the output data is used as the calibration search object, and the particle swarm optimization algorithm is used for optimization search. Here, each set of output data and the corresponding output interpolation code can be iterated 50 times, and each set includes 56 output interpolation codes.

[0068] (3) After the optimization search is completed, the M weights obtained by iteration are averaged as the optimal weight output, which can avoid interference from non-ideal factors such as noise.

[0069] After that, M groups of data are received again, and the optimal weight obtained from the previous optimization search is used as the center value of the initial search range of this optimization search. The initial search range is redefined, and then the optimal weight of this optimization search is determined through multiple iterations until the ADC output data is processed.

[0070] In this example, the optimal weights obtained through optimization search are: [0.499400 0.498688 0.2503080.125632 0.062654 0.031070 0.015628 0.007919 0.003847 0.001993 0.0010230.000477 0.00021 0.00011 0.00005].

[0071] Furthermore, the performance comparison is performed by performing fast Fourier transform on the ADC output data before and after calibration, as shown in Figure 2. Figure 3-4 As shown, the quantization noise before calibration is about 20dB higher than that after calibration. The specific performance index comparison is shown in Table 2:

[0072] Table 2

[0073] ENOB SNDR SFDR THD Before calibration 8.48 bits 52.78dB 68.01dB -65.77dB After calibration 11.93 bits 73.60dB 91.33dB -85.52dB

[0074] It can be seen that the ADC's ENOB (Effective Number of Bits) was 8.48 bits before calibration and 11.93 bits after calibration, an improvement of 3.45 bits. The ADC's SNDR (Signal-to-Noise-and-Distortion Ratio) was 52.78 dB before calibration and 73.60 dB after calibration, an improvement of 20.82 dB. The ADC's SFDR (Spurious-free Dynamic Range) was 68.01 dB before calibration and 91.33 dB after calibration, an improvement of 23.32 dB. The THD (Total Harmonic Distortion) was -65.77 dB before calibration and -85.52 dB after calibration, an improvement of 19.75 dB. Obviously, the intelligent optimization algorithm method provided by the present invention can perform effective digital background calibration on the SAR ADC, and improve various performances.

[0075] It can be seen from the above embodiments that the beneficial effects of the present invention are:

[0076] The present invention provides an ADC digital background calibration method based on an intelligent optimization algorithm. By using an interpolation particle swarm optimization algorithm as a framework to optimize and search for a target optimization function, with the goal of minimizing the mean square error between each set of output data and its corresponding interpolated output code, the optimal weight of the optimization search is obtained, thereby improving the accuracy of a successive approximation ADC and simply and efficiently achieving calibration of the ADC digital background.

[0077] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0078] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification.

[0079] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. For those skilled in the art of the present invention, without departing from the concept of the present invention, several simple deductions or substitutions can be made, which should be considered to fall within the scope of protection of the present invention.

Claims

1. An ADC digital background calibration method based on an intelligent optimization algorithm, characterized in that: include: After receiving M groups of output data sent by the analog-to-digital converter, construct M groups of interpolation output codes using the Lagrange interpolation algorithm; A target optimization function is constructed based on M groups of output data and M groups of interpolation output codes, and the target optimization function is optimized and searched using a particle swarm optimization algorithm to obtain M groups of weights so that the mean square error between each group of output data and its corresponding interpolation output code is minimized; the target optimization function is: in, represents the weight to be optimized, represents the actual output of the j-th iteration of each set of output data, represents the output of the interpolation, Indicates the preset number of iterations when optimizing the search. and The dimensions are all 1×N, The dimension is N×1, where N represents the digital output bit width of the analog-to-digital converter; The average value of the M groups of weights is output as the optimal weight for this optimization search; After the step of outputting the average value of the M groups of weights as the optimal weight of this optimization search, the method further includes: Returning to the step of receiving M groups of output data sent by the analog-to-digital converter, and determining an initial search range of the particle swarm optimization algorithm for the next optimization search of each group of output data and the interpolation output code according to the average value; Wherein, the initial search range of the particle swarm optimization algorithm during the next optimization search takes the average value as the center value.

2. The ADC digital background calibration method based on the intelligent optimization algorithm according to claim 1, characterized in that: After receiving M groups of output data sent by the analog-to-digital converter, the step of constructing M groups of interpolation output codes using a Lagrange interpolation algorithm includes: receiving M groups of output data sent by the analog-to-digital converter; Determine the order of the Lagrange interpolation algorithm used for fitting interpolation: , where N represents the digital output bit width of the analog-to-digital converter is N bits, Indicates rounding down; For each set of output data, an interpolation output code is constructed according to the order of the fitting interpolation using a Lagrange interpolation algorithm to obtain M sets of interpolation output codes corresponding to the M sets of output data.

3. The ADC digital background calibration method based on the intelligent optimization algorithm according to claim 2, characterized in that: Before the step of constructing an interpolation output code according to the order of the fitting interpolation using the Lagrange interpolation algorithm for each set of the output data, the method further includes: Capacitance mismatch in the analog-to-digital converter is estimated, and an initial search range of the particle swarm optimization algorithm is set when optimizing each set of output data.

Images