A bandwidth compensation method based on amplitude differentiation

Abstract

The application discloses a bandwidth compensation method based on amplitude differentiation, sweeps all channels at any two amplitude levels to obtain waveform amplitude values at various frequency points, then divides compensation filter groups by calculating amplitude differentiation degrees of various acquisition channels, designs a compensation filter for any one acquisition channel in the same group, and other channels in the group share the compensation filter, and finally a host computer sends filter coefficients to an FPGA end, and the FPGA end uses the coefficients to perform filter operation, thereby completing bandwidth compensation.

Description

Technical Field

[0001] This invention belongs to the field of high-speed data acquisition technology, and more specifically, relates to a bandwidth compensation method based on amplitude differentiation. Background Technology

[0002] In the design of high-speed data acquisition systems, exemplified by digital oscilloscopes, bandwidth compensation technology is a research hotspot and a key challenge requiring breakthroughs. Bandwidth compensation aims to increase the overall measurement bandwidth of the acquisition system, typically using the -3dB bandwidth limit as a standard. Bandwidth compensation technology aims to maximize the -3dB bandwidth frequency of the acquisition system and flatten the amplitude-frequency response curve of the acquisition channel. The analog bandwidth of the core chip is a major bottleneck in achieving bandwidth breakthroughs in acquisition systems. The limitations in developing high-bandwidth chips increase the difficulty of designing high-bandwidth acquisition channels, significantly restricting their performance in amplitude-frequency response, linearity, and sensitivity. To overcome the limitations of chip bandwidth to the greatest extent possible, bandwidth compensation for analog channels using digital signal processing technology is an industry-recognized approach. Bandwidth compensation typically utilizes digital filtering techniques to improve the bandwidth and frequency response flatness of the acquisition system, making it a crucial method for enhancing system performance.

[0003] The existing solution for bandwidth compensation of the acquisition path is to increase the acquisition bandwidth by filtering the input waveform of a single acquisition path with a digital filter. However, in data acquisition systems represented by digital oscilloscopes, there are multiple acquisition channels, each with multiple amplitude levels. This means that the entire acquisition system has multiple acquisition paths, and each acquisition path has inconsistent amplitude and frequency responses. The filtering compensation design of a single channel cannot be reused. Therefore, the traditional single-path digital filtering compensation technology can no longer meet the bandwidth compensation requirements of the entire acquisition system. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a bandwidth compensation method based on amplitude differentiation, which effectively integrates bandwidth compensation with amplitude levels and improves the sampling performance of the data acquisition system.

[0005] To achieve the above-mentioned objective, the present invention provides a bandwidth compensation method based on amplitude differentiation, characterized by comprising the following steps:

[0006] (1) Initialize the TIADC acquisition system;

[0007] Suppose that the TIADC acquisition system has N acquisition channels; select any two amplitude ranges in the TIADC acquisition system, denoted as amplitude range a and amplitude range b;

[0008] (2) Obtain the amplitude of each acquisition channel of the TIADC acquisition system at each frequency point;

[0009] (2.1) Obtain the amplitude of each acquisition channel of the TIADC acquisition system at each frequency point under each amplitude level;

[0010] (2.1.1) Use a sine wave to start scanning the frequency of each acquisition channel with a fixed frequency step. The amplitude is set so that the sine wave can be displayed on the oscilloscope screen at more than six divisions at the initial frequency, which is the calibration requirement. Stop scanning the frequency when the frequency reaches the set maximum frequency point. A total of M frequency points are scanned. Then, the waveform data of each frequency point scanned in each channel is uploaded to the host computer.

[0011] (2.1.2) The host computer processes the waveform data of each frequency point in each channel to obtain the waveform amplitude value of each frequency point in each channel. Among them, when the amplitude range is a, the waveform amplitude value of the m-th frequency point in the n-th acquisition channel is recorded as... When the amplitude level is b, the waveform amplitude value at the m-th frequency point in the n-th acquisition channel is recorded as... ω m Let m represent the angular frequency of the m-th frequency point, where m = 1, 2, ..., M, and n = 1, 2, ..., N;

[0012] (3) Calculate the amplitude difference of each acquisition channel;

[0013] (3.1) Calculate the amplitude difference ΔD between the two amplitude levels at each frequency point in the nth acquisition channel. n (ω m );

[0014]

[0015] (3.2) Calculate the amplitude difference F of the nth acquisition channel. n ;

[0016] F n =max(ΔD) n (ω m ))

[0017] (4) Divide the compensation coefficient reuse group;

[0018] Set the channel amplitude difference threshold F max ; Iterate through the amplitude difference of each acquisition channel and calculate the amplitude difference F. n Less than F max The acquisition channels are divided into one group, and the remaining acquisition channels are divided into another group;

[0019] (5) Obtain the frequency response function of each group and design a compensation filter;

[0020] (5.1) Within the same group, the waveform amplitude value S(ω1) of the first frequency point of a certain acquisition channel at any amplitude level is used as the quantization reference for frequency response acquisition, and S(ω m The amplitude of the waveform at each frequency point within the sweep range is represented by ), and the frequency response H(jω) at each frequency point in the acquisition channel is then calculated using the following formula. m );

[0021]

[0022] Where, ω m This represents the digital angular frequency of the m-th frequency point;

[0023] Then, the host computer performs curve fitting on the frequency response at each frequency point to obtain the frequency response curve H(jω) of the acquisition channel, where ω represents the digital angular frequency.

[0024] (5.2) Design the function F(jω) of the compensation filter based on the frequency response H(jω) of the acquisition path;

[0025]

[0026] (6) The host computer performs bandwidth compensation on the input data;

[0027] (6.1) The host computer generates the compensation filter coefficients according to the function F(jω) of the compensation filter using the Krylov subspace frequency domain compensation filter design algorithm, and then sends the compensation filter coefficients to the FPGA for caching.

[0028] (6.2) After the TIADC acquisition system acquires data, the filtering module in the FPGA performs rate matching operation on the acquired data to make the compensation filter coefficients and the acquired data have the same acquisition rate. Then the filtering module uses the compensation filter coefficients to filter the acquired data to obtain the waveform data after bandwidth compensation, thus completing the bandwidth compensation of the acquisition channel.

[0029] The objective of this invention is achieved as follows:

[0030] This invention is based on a bandwidth compensation method with amplitude differences. It performs frequency sweeping operations on all channels at any two amplitude levels to obtain waveform amplitude values ​​at each frequency point. Then, it divides the compensation coefficient groups by calculating the amplitude difference of each acquisition channel. For any acquisition channel in the same group, a compensation filter is designed, and other channels in the same group share the compensation filter. Finally, the host computer sends the filter coefficients to the FPGA, and the FPGA uses the coefficients to perform filtering operations to complete the bandwidth compensation.

[0031] Meanwhile, the bandwidth compensation method based on amplitude differentiation of the present invention also has the following beneficial effects:

[0032] (1) In the overall implementation of bandwidth compensation, quantitative indicators such as amplitude difference degree are proposed, and amplitude levels with small amplitude difference degree are divided into a group.

[0033] (2) By sweeping the frequency only for one gear in the same group, a group can share a set of filter coefficients, which makes the bandwidth compensation process faster and makes up for the time-consuming disadvantage of traditional bandwidth compensation methods.

[0034] (3) The control logic of the channel and amplitude was implemented on the host computer, and the filtering coefficients of different levels of different channels were accurately sent. Attached Figure Description

[0035] Figure 1 This is a flowchart of the bandwidth compensation method based on amplitude differentiation of the present invention;

[0036] Figure 2 This is a diagram illustrating bandwidth sweep frequency;

[0037] Figure 3 This is a schematic diagram illustrating the quantitative analysis of frequency response differences;

[0038] Figure 4 This is a schematic diagram illustrating the acquisition of the target filtering function;

[0039] Figure 5 This is a diagram of the filter coefficient generation implementation structure; Detailed Implementation

[0040] The specific embodiments of the present invention will now be described with reference to the accompanying drawings to enable those skilled in the art to better understand the invention. It should be particularly noted that in the following description, detailed descriptions of known functions and designs that might obscure the main content of the invention will be omitted here.

[0041] Example

[0042] Figure 1 This is a flowchart of the bandwidth compensation method based on amplitude differentiation of the present invention.

[0043] In this embodiment, as Figure 1 As shown, the present invention provides a bandwidth compensation method based on amplitude differentiation, comprising the following steps:

[0044] S1. Initialize the TIADC acquisition system;

[0045] Suppose that the TIADC acquisition system has a total of 8 acquisition channels with a bandwidth of 8GHz; the TIADC acquisition system has multiple amplitude levels, such as 10mV / div, 20mV / div, ... 1V / div, etc. We randomly select two amplitude levels, denoted as amplitude level a and amplitude level b.

[0046] In this embodiment, the amplitude level grouping of the 8GHz channel is shown in Table 1;

[0047] Group 1 10mV / div, 20mV / div, 50mV / div Highly similar Group 2 100mV / div, 200mV / div Highly similar Group 3 500mV / div, 1V / div resemblance

[0048] Table 1

[0049] S2. Obtain the amplitude of each acquisition channel of the TIADC acquisition system at each frequency point;

[0050] S2.1 Obtain the amplitude of each acquisition channel of the TIADC acquisition system at each frequency point under each amplitude level;

[0051] S2.1.1. Sweep the frequency of each acquisition channel using a sine wave with an initial frequency of 100MHz in 100MHz increments. The amplitude should be set so that the sine wave displays at least six bars on the oscilloscope screen at the initial frequency, meeting the calibration requirement. To better compensate for bandwidth, the sweep cutoff frequency should be set after the bandwidth. For example, set the sweep cutoff frequency of the 8GHz bandwidth channel to 8.2GHz. Figure 2 As shown, a total of 82 frequency points are scanned, and then the waveform data of each frequency point scanned in each channel is uploaded to the host computer;

[0052] S2.1.2 The host computer processes the waveform data of each frequency point in each channel to obtain the waveform amplitude value at each frequency point in each channel. Specifically, when the amplitude level is 'a', the waveform amplitude value at the m-th frequency point in the n-th acquisition channel is recorded as... When the amplitude level is b, the waveform amplitude value at the m-th frequency point in the n-th acquisition channel is recorded as... ω m Let represent the angular frequency of the m-th frequency point, where m = 1, 2, ..., 82 and n = 1, 2, ..., 8;

[0053] S3. Calculate the amplitude difference of each acquisition channel;

[0054] S3.1 Calculate the amplitude difference ΔD between two amplitude levels at each frequency point in the nth acquisition channel. n (ω m );

[0055]

[0056] S3.2 Calculate the amplitude difference F of the nth acquisition channel. n ;

[0057] F n =max(ΔD) n (ω m ))

[0058] S4. Divide the compensation coefficient reuse group;

[0059] Set the channel amplitude difference threshold F max ; Iterate through the amplitude difference of each acquisition channel and calculate the amplitude difference F. n Less than F max The acquisition channels are divided into one group, and the remaining acquisition channels are divided into another group;

[0060] In this embodiment, the TIADC acquisition system has 8 acquisition channels, each with multiple vertical ranges. Sweeping the frequency for each vertical range of each channel is a time-consuming task. Therefore, we group vertical ranges with similar amplitudes within the same acquisition channel into a single group based on amplitude differences. Thus, in step S5, vertical ranges within the same group reuse a set of compensation coefficients, thereby reducing the workload of system bandwidth compensation. A schematic diagram of the quantitative analysis of frequency response differences is shown below. Figure 3 As shown;

[0061] S5. Obtain the frequency response function for each group and design a compensation filter;

[0062] S5.1 Within the same group, the waveform amplitude value S(ω1) of the first frequency point of a certain acquisition channel at any amplitude level is used as the quantization reference for frequency response acquisition, and S(ω m The amplitude of the waveform at each frequency point within the sweep range is represented by ), and the frequency response H(jω) at each frequency point in the acquisition channel is then calculated using the following formula. m );

[0063]

[0064] Where, ω m This represents the digital angular frequency of the m-th frequency point;

[0065] Then, the host computer performs curve fitting on the frequency response at each frequency point to obtain the frequency response curve H(jω) of the acquisition channel, where ω represents the digital angular frequency.

[0066] S5.2, such as Figure 4 As shown, the function F(jω) of the compensation filter is obtained based on the frequency response H(jω) of the acquisition path;

[0067]

[0068] S6. The host computer performs bandwidth compensation on the input data;

[0069] S6.1, such as Figure 5 As shown, the host computer generates the compensation filter coefficients using the Krylov subspace frequency domain compensation filter design algorithm based on the function F(jω) of the compensation filter, and then sends the compensation filter coefficients to the FPGA for caching.

[0070] S6.2 After the TIADC acquisition system acquires data, the filtering module in the FPGA performs a rate matching operation on the acquired data to ensure that the compensation filter coefficients and the acquired data have the same acquisition rate. Then, the filtering module uses the compensation filter coefficients to filter the acquired data to obtain the waveform data after bandwidth compensation, thus completing the bandwidth compensation of the acquisition channel.

[0071] Although the illustrative specific embodiments of the present invention have been described above to enable those skilled in the art to understand the invention, it should be understood that the invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the invention as defined and determined by the appended claims, and all inventions utilizing the concept of the present invention are protected.

Claims

1. A bandwidth compensation method based on amplitude differentiation, characterized in that, Includes the following steps: (1) Initialize the TIADC acquisition system; Assume the TIADC acquisition system has N acquisition channels; In the TIADC acquisition system, select any two amplitude ranges, denoted as amplitude range a and amplitude range b. (2) Obtain the amplitude of each acquisition channel of the TIADC acquisition system at each frequency point; (2.1) Obtain the amplitude of each acquisition channel of the TIADC acquisition system at each frequency point under each amplitude level; (2.1.1) Use a sine wave to start scanning the frequency of each acquisition channel with a fixed frequency step. The amplitude is set so that the sine wave can be displayed on the oscilloscope screen at more than six divisions at the initial frequency, which is the calibration requirement. Stop scanning the frequency when the frequency reaches the set maximum frequency point. A total of M frequency points are scanned. Then, the waveform data of each frequency point scanned in each channel is uploaded to the host computer. (2.1.2) The host computer processes the waveform data of each frequency point in each channel to obtain the waveform amplitude value of each frequency point in each channel. Among them, when the amplitude range is a, the waveform amplitude value of the m-th frequency point in the n-th acquisition channel is recorded as... When the amplitude level is b, the waveform amplitude value at the m-th frequency point in the n-th acquisition channel is recorded as... ω m Let m represent the angular frequency of the m-th frequency point, where m = 1, 2, ..., M, and n = 1, 2, ..., N; (3) Calculate the amplitude difference of each acquisition channel; (3.1) Calculate the amplitude difference ΔD between the two amplitude levels at each frequency point in the nth acquisition channel. n (ω m ); (3.2) Calculate the amplitude difference F of the nth acquisition channel. n ; F n =max(ΔD n (oh m )) (4) Divide the compensation coefficient reuse group; Set the channel amplitude difference threshold F max ; Iterate through the amplitude difference of each acquisition channel and calculate the amplitude difference F. n Less than F max The acquisition channels are divided into one group, and the remaining acquisition channels are divided into another group; (5) Obtain the frequency response function of each group and design a compensation filter; (5.1) Within the same group, the waveform amplitude value S(ω1) of the first frequency point of a certain acquisition channel at any amplitude level is used as the quantization reference for frequency response acquisition, and S(ω m The amplitude of the waveform at each frequency point within the sweep range is represented by ), and the frequency response H(jω) at each frequency point in the acquisition channel is then calculated using the following formula. m ); Where, ω m This represents the digital angular frequency of the m-th frequency point; Then, the host computer performs curve fitting on the frequency response at each frequency point to obtain the frequency response curve H(jω) of the acquisition channel, where ω represents the digital angular frequency. (5.2) Design the function F(jω) of the compensation filter based on the frequency response H(jω) of the acquisition path; (6) The host computer performs bandwidth compensation on the input data; (6.1) The host computer generates the compensation filter coefficients according to the function F(jω) of the compensation filter using the Krylov subspace frequency domain compensation filter design algorithm, and then sends the compensation filter coefficients to the FPGA for caching. (6.2) After the TIADC acquisition system acquires data, the filtering module in the FPGA performs rate matching operation on the acquired data to make the compensation filter coefficients and the acquired data have the same acquisition rate. Then the filtering module uses the compensation filter coefficients to filter the acquired data to obtain the waveform data after bandwidth compensation, thus completing the bandwidth compensation of the acquisition channel.

Images