Fundamental wave signal extraction system and CIC filter

By combining multi-order differentializers in the fundamental signal extraction system and controlling the differentializer input using the switching module, the number of differentiators is reduced, and the problems of complex structure and large area of ​​the CIC filter are solved, and the accurate extraction of the fundamental signal is achieved. Frequency band compensation simplifies circuit design.

CN119030499APending Publication Date: 2024-11-26SPL ELECTRONICS TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411192376.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

In the existing fundamental signal extraction system, CIC filters require N cascaded integrators and N cascaded differentiators, resulting in complex calculation circuits and large area occupancy.

Method used

Adopt an improved CIC filter structure, in which multi-order differentials are combined into one, and the differential input is controlled by switching modules, reducing the number of differentials, and only one subtractor cycles N-1 times, simplifying the circuit structure. And combine the compensation filter to perform band compensation on the CIC filter output.

Benefits of technology

The area of ​​the CIC filter is effectively reduced, the circuit structure is simplified, and the accuracy of fundamental signal extraction is achieved, and the passband compensation and stopband attenuation are completed through the compensation filter, reducing design complexity.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119030499A_ABST
    Figure CN119030499A_ABST
Patent Text Reader

Abstract

The invention relates to a fundamental wave signal extraction system and a CIC filter, and belongs to the technical field of digital signal processing. According to the CIC filter, multi-order differentiators in an existing CIC filter are combined into one, one-time differentiator operation is achieved through subtractor cycle operation, namely, the effect that one subtractor cycles N times to replace N subtractors is achieved, input of the differentiators is controlled through the switching module, and the functions of the multiple differentiators are achieved through multiplexing of one differentiator. Therefore, by reducing the number of differentiators, the circuit structure of the CIC filter can be simplified, and the area of the CIC filter can be effectively reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a fundamental wave signal extraction system and a CIC filter, belonging to the technical field of digital signal processing. Background Art

[0002] The widespread use of nonlinear loads, especially power electronic equipment, in power systems has led to serious harmonic pollution in the power grid, which has caused voltage and current waveform distortion and endangered the safe and efficient operation of other linear devices in the grid. In order to reduce harmonic pollution in the power grid and ensure the fairness of power supply and use, it is necessary to measure the fundamental power and full-wave power of the load under harmonic conditions separately.

[0003] The core of calculating fundamental electric energy under harmonic conditions is how to accurately extract the fundamental signal from the full-wave signal, that is, to weaken the harmonic signals above the second order. The commonly used technology is to filter out the harmonic signal components through a low-pass filter. At present, most mainstream metering chips have the function of high-order harmonic (up to 61 times) analysis, so the sampling frequency of AD reaches above 12.8KHz. To obtain a 50Hz fundamental signal, a narrowband low-pass filter needs to be designed. For the design of narrowband low-pass filters, if FIR filters are used for design, the number of FIR taps is large and the module area is large; if IIR filters are used for design, the fundamental signal is lost after fixed-point implementation.

[0004] The Chinese patent application document with application publication number CN117233460A discloses a digital hardware system for high-precision fundamental wave measurement under harmonic and frequency deviation conditions. The scheme first uses a CIC filter to sample the full-wave signal, then uses an IIR compensation filter to compensate for the passband attenuation, and then uses a low-pass filter to increase the attenuation of harmonics above the second order. This method requires more filters to be designed and the design is more complicated. In addition, the CIC filter implemented by the scheme is composed of N identical integrators and N identical differentiators in cascade. The process of CIC filter sampling is: the input data first enters N cascaded integrators, and then is sampled 32 times. The sampled data passes through N cascaded differentiators in sequence. For a three-phase four-wire channel, there are usually 7 channels inside the metering chip, and each channel needs to pass through a CIC filter. The CIC filter of each channel contains N adders and N subtractors. The filter contains more calculation processes, which is reflected in the actual circuit as a complex circuit, and at this time, it needs to occupy a larger area. Summary of the invention

[0005] The purpose of the present invention is to provide a fundamental wave signal extraction system and a CIC filter to solve the problem that the CIC filter used in the current fundamental wave signal extraction system requires N cascaded integrators and N cascaded differentiators, which leads to a complex calculation circuit and a large area occupied by the circuit on the chip.

[0006] In order to solve the above technical problems, the present invention provides a fundamental signal extraction system, which includes a CIC filter and a compensation filter. The CIC filter includes N integrators, 1 differentiator, 1 switching module and N buffers, N is the order of the CIC filter, N is determined by the down sampling multiple R, and one down sampling process consists of R counting cycles; N integrators are cascaded, and the switching module selects at most one from the output of the Nth integrator and the buffer connected to the differentiator output according to the counting cycle of the down sampling process and divides it into two paths, one path is directly output to the differentiator as the input of the differentiator, and the other path is cascaded through N-1 buffers as another input of the differentiator, and the differentiator is used to complete N-1 differential operations within the counting cycle of a down sampling process; the compensation filter is used to compensate for the useful frequency band of the CIC filter output data.

[0007] Furthermore, the order N of the CIC filter needs to satisfy: N is less than R / 2+1, where R is the downsampling multiple.

[0008] Furthermore, when the switching module is switching, the output of the Nth integrator is selected as the input of the differentiator in the 0th counting cycle of a down sampling process, and the buffer connected to the differentiator output is selected as the input of the differentiator in the N-2 counting cycles between the 1st and R-1th counting cycles of a down sampling process, and the differentiator does not work in the remaining counting cycles.

[0009] Furthermore, the transfer function of the CIC filter is:

[0010]

[0011] Where N is the order of the CIC filter, R is the downsampling factor, (1-z -R ) N It is the comb filter part.

[0012] Furthermore, the transfer function amplitude response H of the CIC filter cic (e jw )| is:

[0013]

[0014] Where N is the order of the CIC filter, R is the downsampling factor, and w is the angular frequency of the input signal.

[0015] Furthermore, the amplitude response H of the compensation filter comp (e jw ) is obtained by taking the inverse of the amplitude response of the CIC filter.

[0016] Furthermore, the extraction system also includes a DC removal filter, and the signal to be extracted input by the first integrator is a full-wave signal that does not contain a DC component after being filtered by the DC removal filter.

[0017] The present invention also provides a CIC filter, which includes N integrators, 1 differentiator, 1 switching module and N buffers, N is the order of the CIC filter, N is determined by the down-sampling multiple R, and one down-sampling process consists of R counting cycles; the N integrators are cascaded, and the switching module selects at most one from the buffer connected to the output of the Nth integrator and the output of the differentiator according to the counting cycle of the down-sampling process and divides it into two paths, one path is directly output to the differentiator as the input of the differentiator, and the other path is cascaded through N-1 buffers as another input of the differentiator, and the differentiator is used to complete N-1 differential operations within the counting cycle of a down-sampling process.

[0018] Furthermore, when the switching module is switching, the output of the Nth integrator is selected as the input of the differentiator in the 0th counting cycle of a down sampling process, and the buffer connected to the differentiator output is selected as the input of the differentiator in the N-2 counting cycles between the 1st and R-1th counting cycles of a down sampling process, and the differentiator does not work in the remaining counting cycles.

[0019] Furthermore, the order N of the CIC filter needs to satisfy: N is less than R / 2+1, where R is the downsampling multiple.

[0020] The beneficial effects of the present invention are as follows: as an improved invention, the present invention synthesizes the multi-order differentiators in the existing CIC filter into one, and uses a subtractor cyclic operation to realize a differentiator operation, that is, one subtractor circulates N times to replace the effect of N subtractors. Due to the reset operation, one subtractor can be removed, and finally only one subtractor needs to circulate N-1 times to obtain the effect of N-1 subtractors. The present invention controls the input of the differentiator by switching the module, so as to realize the effect of multiple differentiators by multiplexing one differentiator. Therefore, the present invention can simplify the circuit structure of the CIC filter by reducing the number of differentiators, so that the area of ​​the CIC filter can be effectively reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a digital hardware system block diagram of the fundamental wave signal extraction system of the present invention;

[0022] Figure 2 It is the principle diagram of the CIC filter adopted by the present invention;

[0023] Figure 3 is a spectrum response diagram in the 0-200 Hz frequency band after compensation in an embodiment of the present invention;

[0024] Figure 4 0-80 Hz frequency band after compensation in an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings.

[0026] The present invention combines the multi-order differentiators in the existing CIC filter into one, and controls the input of the differentiator through a switching module, so as to realize the function of multiple differentiators through the reuse of one differentiator. Therefore, the present invention can simplify the circuit structure of the CIC filter by reducing the number of differentiators, so that the area of ​​the CIC filter can be effectively reduced. In addition, the present invention is to obtain the spectrum response of the compensation filter by deriving the spectrum response of the CIC filter, and then generate the compensation filter coefficient according to the spectrum response of the compensation filter using the fir2 function. The fir2 function can set the passband frequency. Within the passband frequency, the spectrum response of the compensation filter and the CIC filter have opposite characteristics, so the passband is compensated, and above the passband frequency, the spectrum response of the compensation filter is gradually attenuated, so it has the function of stopband attenuation. The present invention can complete the two functions of passband compensation and stopband attenuation with only one compensation filter, without increasing the stopband attenuation, and the design is relatively simple.

[0027] Embodiment of fundamental signal extraction system

[0028] The fundamental wave signal extraction system of this embodiment can be applied to the electric energy metering chip, such as Figure 1 As shown, it includes a DC removal filter, a CIC filter and a compensation filter. Since the analog-to-digital converter used in the metering chip will generate a certain amount of DC, and the DC will affect the subsequent fundamental electric energy, the DC removal filter can be used to first filter out the DC component in the collected signal to obtain a full-wave signal without a DC component; the CIC filter is used to perform down-sampling filtering on the obtained full-wave signal, and multiple-order differentiators are combined into one differentiator, which simplifies the structure of the CIC filter. The compensation filter is used to compensate for the useful frequency band of the data output by the CIC filter.

[0029] The transfer function H(z) of the DC filter is:

[0030]

[0031] Where a is the filter coefficient. The larger a is, the weaker the DC removal effect is and the faster the DC removal convergence process is. The smaller a is, the stronger the DC removal effect is and the slower the DC removal convergence process is.

[0032] The transfer function of the CIC filter is:

[0033]

[0034] Where: N is the order of the CIC filter, R is the downsampling multiple, and the comb filter part is (1-z -R ) N , H cic (z) Amplitude response |H cic (e jw )| is:

[0035]

[0036] where w is the angular frequency.

[0037] The magnitude response of the compensation filter H comp (e jw ) is the inverse of the amplitude response of the CIC filter:

[0038]

[0039] The CIC filter uses a multiplexed CIC filter, including N integrators, 1 differentiator, 1 switching module and N buffers, where N is the order of the CIC filter; the N integrators are connected in sequence, the output of the previous integrator is the input of the next integrator, the input of the first integrator is the signal to be extracted, the output of the Nth integrator is used as an input of the switching module, the output of the differentiator is used as another input of the switching module after passing through a buffer, and the switching module selects at most one input from the two inputs according to the counting cycle of the down sampling process and divides it into two paths, one path is directly output to the differentiator as the input of the differentiator, and the other path is cascaded through N-1 buffers as another input of the differentiator. Due to the reset operation, one subtractor can be removed, and finally only one subtractor needs to be cycled N-1 times to obtain the effect of N-1 subtractors. In addition, the down sampling multiple R must be at least greater than 2*(N-1) to ensure the effectiveness of the structure proposed in the patent, where N is the order.

[0040] The following description is made by taking the original sampling frequency of the signal to be extracted as 12.8KHz as an example. The order of the CIC filter in this embodiment is 4, the downsampling multiple is 32 times, and there are 32 counting cycles in total. Figure 2As shown, it includes 4 integrators, namely sum1, sum2, sum3 and sum4, 4 buffers, namely temp1, temp2, temp3 and temp4, the switching module adopts a conversion switch EN_MUX, and there is also a differentiator, wherein sum1, sum2, sum3 and sum4 are connected in sequence, that is, the input of sum1 is used to receive the full-wave signal after de-straightening, the output of sum1 is the input of sum2, the output of sum2 is the input of sum3, the output of sum3 is the input of sum4, the output of sum4 is one input of the conversion switch EN_MUX, and the output of the conversion switch EN_MUX is divided into two paths, one path is directly used as the input of the differentiator, and the other path is used as another input of the differentiator after passing through temp1, temp2 and temp3, and the output of the differentiator is used as one input of the conversion switch EN_MUX after passing through temp4. After a down-sampling filtering process is completed, the value in temp4 is the down-sampling filtering result of the CIC filter. The working process of the CIC filter is described below. To facilitate the description of the process, a counter is used to count the counting cycle of a down-sampling filtering process, and the counter starts counting from 0.

[0041] The sampling multiple in this embodiment is 32 times, so a result is output every 32 points. From the 0th to the 31st point, the integrator needs to perform operations at each point, while the differentiator only needs to perform operations once. Since only 3 subtractor operations need to be completed in the 0-31 time, some operations are the same and some operations are different. In this embodiment, the 0th point performs a subtraction operation, and the final result should be output. The 1st point stops working, the 2nd point performs a subtraction operation, the 3rd point stops working, and the 4th point performs a subtraction operation. So far, 3 subtraction operations have been completed, so the subsequent counting cycle differentiator part stops working and keeps the register state value. In addition, in this embodiment, the first 5 times stop working every other point. This does not have to be every other point, and can also be every two or three points, as long as it is guaranteed that 3 subtraction operations can be circulated in the 0-31 cycle. The following is a detailed description of its working process.

[0042] When the counter is 0, that is, the 0th counting cycle, the input full-wave data enters the integrator, and the integrators sum1, sum2, sum3 and sum4 work at the same time. The output of integrator sum1 is used as the input of sum2, the output of integrator sum2 is used as the input of sum3, and the output of integrator sum3 is used as the input of sum4. The conversion switch EN_MUX is 1, and the output of sum4 is selected as the input of the differentiator. At the same time, the value in the buffer in sum4 is set to 0, and the input value of the differentiator is subtracted from the value of buffer temp3. The value of buffer temp2 is assigned to temp3, the value of buffer temp1 is assigned to temp2, the input value of the differentiator is assigned to temp1, and the output of the differentiator is assigned to buffer temp4 and output, and the counter is incremented by 1 to complete the operation within a counting cycle;

[0043] When the counter is 1, that is, the first counting cycle, the input data enters the integrator, and the integrators sum1, sum2, sum3 and sum4 work at the same time. The output of the integrator sum1 is used as the input of sum2, the output of the integrator sum2 is used as the input of sum3, and the output of the integrator sum3 is used as the input of sum4. After the output of the integrator sum4 is completed, the conversion switch EN_MUX is 0, the differentiator does not work, and no output is made. The counter is incremented by 1, completing the operation within a counting cycle.

[0044] When the counter is 2, that is, the second counting cycle, the input data enters the integrator, and the integrators sum1, sum2, sum3 and sum4 work at the same time. The output of integrator sum1 is used as the input of sum2, the output of integrator sum2 is used as the input of sum3, and the output of integrator sum3 is used as the input of sum4. After the output of integrator sum4 is completed, the conversion switch EN_MUX is 0, and the buffer temp4 is used as the input of the differentiator. The input value of the differentiator is subtracted from the value of buffer temp3, and the value of buffer temp2 is assigned to temp3, the value of buffer temp1 is assigned to temp2, and the input value of the differentiator is assigned to temp1. The counter is incremented by 1, and the output of the differentiator is assigned to buffer temp4. No output is performed, and the counter is incremented by 1 to complete the operation within a counting cycle.

[0045] When the counter is 3, that is, the third counting cycle, the input data enters the integrator, and the integrators sum1, sum2, sum3 and sum4 work at the same time. The output of integrator sum1 is used as the input of sum2, the output of integrator sum2 is used as the input of sum3, and the output of integrator sum3 is used as the input of sum4. After the output of integrator sum4 is completed, the conversion switch EN_MUX is 0, the differentiator does not work, and no output is made. The counter is increased by 1, and the operation within a counting cycle is completed;

[0046] When the counter is 4, the input data enters the integrator, and the integrators sum1, sum2, sum3 and sum4 work at the same time. The output of integrator sum1 is used as the input of sum2, the output of integrator sum2 is used as the input of sum3, and the output of integrator sum3 is used as the input of sum4. After the output of integrator sum4 is completed, the conversion switch EN_MUX is 0, and the buffer temp4 is used as the input of the differentiator. The input value of the differentiator is subtracted from the value of buffer temp3, and the value of buffer temp2 is assigned to temp3, the value of buffer temp1 is assigned to temp2, and the input value of the differentiator is assigned to temp1. The counter is incremented by 1, and the output of the differentiator is assigned to buffer temp4. No output is performed, and the counter is incremented by 1 to complete the operation within a counting cycle.

[0047] When the counter is 5, the input data enters the integrator, and the integrators sum1, sum2, sum3 and sum4 work at the same time. The output of integrator sum1 is used as the input of sum2, the output of integrator sum2 is used as the input of sum3, and the output of integrator sum3 is used as the input of sum4. After the output of integrator sum4 is completed, the conversion switch EN_MUX is 0, the differentiator does not work, does not output, and the counter is increased by 1, completing the operation within a counting cycle;

[0048] When the counter is 6, 7, 8, ..., 30, 31, the operation is the same as when the counter is 5;

[0049] When the counter is 32, the input data enters the integrator, and the integrators sum1, sum2, sum3 and sum4 work at the same time. The output of integrator sum1 is used as the input of sum2, the output of integrator sum2 is used as the input of sum3, and the output of integrator sum3 is used as the input of sum4. After the output of integrator sum4 is completed, the conversion switch EN_MUX is 1, and the output of sum4 is selected as the input of the differentiator. At the same time, the value in the buffer in sum4 is set to 0, and the input value of the differentiator is subtracted from the value of buffer temp3. The value of buffer temp2 is assigned to temp3, the value of buffer temp1 is assigned to temp2, the input value of the differentiator is assigned to temp1, and the output of the differentiator is assigned to buffer temp4 and output, the counter is increased by 1, and the operation within a counting cycle is completed;

[0050] When the counter is 33, the operation is the same as when the counter is 1;

[0051] When the counter is 34, the operation is the same as when the counter is 2;

[0052] When the counter is 35, the operation is the same as when the calculator is 3;

[0053] When the counter is 36, the operation is the same as when the counter is 4;

[0054] When the counter is 37, the operation is the same as when the counter is 5;

[0055] When the counter is 38, 39, 40, .... 62, 63, the operation is the same as when the counter is 37;

[0056] When the counter is 64, the input data enters the integrator, and integrators sum1, sum2, sum3 and sum4 work at the same time. The output of integrator sum1 is used as the input of sum2, the output of integrator sum2 is used as the input of sum3, and the output of integrator sum3 is used as the input of sum4. After the output of integrator sum4 is completed, the conversion switch EN_MUX is 1, and the output of sum4 is selected as the input of the differentiator. At the same time, the value in the buffer in sum4 is set to 0, and the input value of the differentiator is subtracted from the value of buffer temp3. The value of buffer temp2 is assigned to temp3, the value of buffer temp1 is assigned to temp2, the input value of the differentiator is assigned to temp1, and the output of the differentiator is assigned to buffer temp4 and output. The counter is increased by 1 to complete the operation within a counting cycle.

[0057] The CIC filter obtains an output every 32 cycles of the counter. When the counter is an integer multiple of 32, the value in the buffer temp4 is the 32-fold downsampling filtering result.

[0058] The above is an introduction to the working process of a CIC filter with an order of 4. The CIC filter can also use other orders according to the frequency of the signal to be extracted. For example, 3rd order, 5th order, etc. If it is 3rd order, there are 3 corresponding integrators, 3 registers in the differentiator part, and 2 cyclic subtraction operations. The specific calculation process is the same; if it is 5th order, there are 5 corresponding integrators, 5 registers in the differentiator part, and 4 cyclic subtraction operations. The specific process is the same and will not be described in detail here.

[0059] The filter result of the above CIC filter can be compensated by a compensation filter. In this embodiment, the spectrum response function of the compensation filter is |H comp (e jw )|, call the fir2(n,f,m) function in matlab to obtain the filter coefficients. The parameter n is the number of filter coefficients, f is the angular frequency (Matlab requires radians when calling, so the angular frequency is used), and m is the spectrum response value corresponding to the angular frequency. In this embodiment, the cutoff frequency of the compensation filter is set to fc=80Hz, and the setting of parameter f can be obtained: f=[f1f2];m settings: The length of the m2 sequence is equal to that of f2, and the value of m2 is all 0, m=[m1m2], n is set to 40, and the coefficients of the compensation filter can be obtained by substituting the parameters n, f and m into the fir2 function.

[0060] After compensation, the spectrum response in the 0-200 Hz frequency band obtained in this embodiment is as follows: Figure 3 As shown, the spectrum response in the 0-80Hz frequency band is as follows Figure 4 As shown. Figure 3 It can be seen that the compensated filter can attenuate harmonics above 100Hz by more than 60dB, greatly reducing the impact of harmonics on the fundamental wave; and from Figure 4 It can be seen that the ripple in the passband of the compensated filter is within 0.02dB.

[0061] It can be seen that the present invention combines multi-order differentiators into one differentiator through multiplexing, which can reduce N differentiators in each channel to one differentiator, reduce the area of ​​the module, and complete the two functions of passband compensation and stopband attenuation with only one compensation filter, which can not only accurately extract the fundamental component, but also simplify the calculation circuit, thereby helping to reduce the area of ​​the chip.

[0062] Example of CIC filter

[0063] The CIC filter of this embodiment adopts a multiplexed CIC filter, including N integrators, 1 differentiator, 1 switching module and N buffers, where N is the order of the CIC filter. The specific structure and working principle have been described in detail in the embodiment of the fundamental signal extraction system and will not be repeated here.

Claims

1. A fundamental signal extraction system, the extraction system comprising a CIC filter and a compensation filter, characterized in that: The CIC filter comprises N integrators, 1 differentiator, 1 switching module and N buffers, wherein N is the order of the CIC filter, and N is determined by the down sampling multiple R, and one down sampling process consists of R counting cycles; the N integrators are cascaded, and the switching module selects at most one of the buffers connected to the output of the Nth integrator and the output of the differentiator according to the counting cycle of the down sampling process, and divides it into two paths, one of which is directly output to the differentiator as the input of the differentiator, and the other path is cascaded through N-1 buffers as another input of the differentiator, and the differentiator is used to complete N-1 differential operations within the counting cycle of one down sampling process; the compensation filter is used to compensate the useful frequency band of the output data of the CIC filter.

2. The fundamental signal extraction system according to claim 1, characterized in that: The order N of the CIC filter needs to satisfy: N is less than R / 2+1, where R is the downsampling multiple.

3. The fundamental wave signal extraction system according to claim 1, characterized in that: When the switching module is switching, the output of the Nth integrator is selected as the input of the differentiator in the 0th counting cycle of a down sampling process, and the buffer connected to the differentiator output is selected as the input of the differentiator in the N-2 counting cycles between the 1st and R-1th counting cycles of a down sampling process, and the differentiator does not work in the remaining counting cycles.

4. The fundamental signal extraction system according to any one of claims 1 to 3, characterized in that: The transfer function of the CIC filter is: Where N is the order of the CIC filter, R is the downsampling multiple, (1-z -R ) N It is the comb filter part.

5. The fundamental wave signal extraction system according to claim 4, characterized in that: The CIC filter transfer function magnitude response H cic (e jw )| is: Where N is the order of the CIC filter, R is the downsampling factor, and w is the angular frequency of the input signal.

6. The fundamental wave signal extraction system according to any one of claims 1 to 3, characterized in that: The magnitude response of the compensation filter H comp (e jw ) is obtained by taking the inverse of the amplitude response of the CIC filter.

7. The fundamental wave signal extraction system according to any one of claims 1 to 3, characterized in that: The extraction system also includes a DC removal filter, and the signal to be extracted input by the first integrator is a full-wave signal that does not contain a DC component after being filtered by the DC removal filter.

8. A CIC filter, characterized in that: The CIC filter comprises N integrators, 1 differentiator, 1 switching module and N buffers, wherein N is the order of the CIC filter, and N is determined by the down sampling multiple R, and one down sampling process consists of R counting cycles; the N integrators are cascaded, and the switching module selects at most one from the buffer connected to the output of the Nth integrator and the output of the differentiator according to the counting cycle of the down sampling process and divides it into two paths, one path is directly output to the differentiator as the input of the differentiator, and the other path is cascaded through N-1 buffers as another input of the differentiator, and the differentiator is used to complete N-1 differential operations within the counting cycle of one down sampling process.

9. The CIC filter according to claim 8, characterized in that: When the switching module is switching, the output of the Nth integrator is selected as the input of the differentiator in the 0th counting cycle of a down sampling process, and the buffer connected to the differentiator output is selected as the input of the differentiator in the N-2 counting cycles between the 1st and R-1th counting cycles of a down sampling process, and the differentiator does not work in the remaining counting cycles.

10. The CIC filter according to claim 8 or 9, characterized in that: The order N of the CIC filter needs to satisfy: N is less than R / 2+1, where R is the downsampling multiple.