A calibration method for ultra-wideband conducted emission direct-coupled probe

Abstract

The present invention discloses a calibration method for an ultra-wideband conducted emission direct coupled probe, which comprises the following steps: performing a dual-port calibration on a vector network analyzer; connecting a through calibration fixture between two ports of the vector network analyzer, and measuring the S matrix of the through calibration fixture; connecting an open-circuit calibration fixture only at the first port of the vector network analyzer, and measuring the input reflection coefficient of the open-circuit calibration fixture; measuring the overall S matrix formed by the open-circuit calibration fixture and the measured coupled probe; performing de-embedding calculation based on the S matrix of the through calibration fixture, the input reflection coefficient of the open-circuit calibration fixture, and the overall S matrix, and obtaining the transfer function of the measured coupled probe; and calibrating the measured coupled probe based on the transfer function of the measured coupled probe. The method can realize the measurement of the transmission characteristics of the probe covering 20GHz, and better meet the demand for conducted emission measurement in the existing integrated circuit field.

Description

Technical Field

[0001] The invention relates to the technical field of microwave measurement, and in particular to a calibration method for an ultra-wideband conducted emission direct coupling probe. Background Art

[0002] Conducted emission (CE) is an electromagnetic phenomenon in which the voltage or current of integrated circuit (IC) chips and electronic devices is transmitted through chip pins, transmission lines and other paths and interferes with other electronic components. The measurement of conducted emission is an important part of the electromagnetic compatibility (EMC) test of integrated circuits and can be used to locate the source of radio frequency (RF) electromagnetic interference. The accuracy of conducted emission measurement is an important guarantee for the design, packaging and integration of integrated circuits and chips. At present, the main methods for conducting emission measurement include near-field magnetic field probe induction measurement method and direct coupling measurement method (DCP). The measurement devices used in these two measurement methods are near-field magnetic field probe and 1Ω / 150Ω probe with impedance network (where 1ΩCE probe is used to measure RF conducted disturbance current and 150ΩCE probe is used to measure RF conducted disturbance voltage). With the continuous expansion of the frequency range of integrated circuits, higher requirements are also put forward for the applicability and accuracy of the calibration method of the conduction measurement device.

[0003] The advantage of the DCP method is that it can locate the disturbance source more directly by direct contact, and can directly measure the disturbance voltage and current values, rather than indirectly obtaining them by electromagnetic field induction like the magnetic field probe method. Due to the simple measurement setup, this method ensures high repeatability and correlation of conducted electromagnetic emission measurements. However, for the calibration method of 1Ω / 150Ω direct coupled probe (DCP), only the calibration method of 1Ω current probe and the structural design of the calibration fixture are given in Appendix A of the standard IEC 61967-4, but the calibration method of 150Ω voltage probe is not given. The calibration method given in the standard is applicable to the test bandwidth of the current probe of 150kHz-1GHz and is suitable for probes of specific structures. With the continuous increase of radio millimeter wave broadband application scenarios, the operating frequency range of the new DCP put into market application has far exceeded 1GHz. The use of the original probe calibration method will result in a decrease in calibration accuracy due to impedance mismatch and phase delay problems, and the impact of the above factors will increase sharply with the increase of calibration frequency. In view of the deficiencies and shortcomings in the above-mentioned 1Ω / 150ΩDCP calibration process, no research results have been proposed so far, so it is necessary to develop a broadband, accurate calibration method that is applicable to both voltage and current probes. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides an ultra-wideband conducted emission direct coupling probe calibration method which solves the problems of insufficient current probe calibration frequency range and lack of voltage probe calibration method in the prior art.

[0005] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0006] A calibration method for an ultra-wideband conducted emission direct coupling probe is provided, comprising the following steps:

[0007] S1. Perform a dual-port calibration on a vector network analyzer, and calibrate the reference end face to the port of the coaxial cable to be measured; wherein the vector network analyzer is a dual-port vector network analyzer;

[0008] S2. Connect a straight-through calibration fixture between two ports of the vector network analyzer and measure the S matrix of the straight-through calibration fixture;

[0009] S3, connecting an open-circuit calibration fixture only to the first port of the vector network analyzer, and measuring the input reflection coefficient of the open-circuit calibration fixture;

[0010] S4, maintaining the connection between the open-circuit calibration fixture and the first port of the vector network analyzer, connecting the ground terminal of the measured coupling probe to the ground terminal of the open-circuit calibration fixture, connecting the input terminal of the measured coupling probe to the output terminal of the open-circuit calibration fixture, connecting the output terminal of the measured coupling probe to the second port of the vector network analyzer, and measuring to obtain an overall S matrix formed by the open-circuit calibration fixture and the measured coupling probe;

[0011] S5, performing de-embedding calculation based on the S matrix of the through calibration fixture, the input reflection coefficient of the open-circuit calibration fixture and the overall S matrix to obtain the transfer function of the coupled probe under test;

[0012] S6. Calibrate the coupled probe under test based on the transfer function of the coupled probe under test.

[0013] Further, the through-calibration fixture includes a 50Ω through-calibration fixture for current probe calibration and a 150Ω through-calibration fixture for voltage probe calibration; wherein:

[0014] The 50Ω through fixture is a microstrip line structure, and the metal conductor line width matches the size of the probe tip to avoid additional distributed inductance and parasitic capacitance;

[0015] The 150Ω through fixture is an impedance gradient microstrip line structure, which converts the 50Ω at the coaxial input end into the 150Ω characteristic impedance on the chip and matches it with the input impedance of the voltage probe to reduce reflections caused by impedance mismatch.

[0016] Furthermore, the length of the metal conductor line of the 50Ω straight-through clamp is 90mm, the width of the metal conductor line of the 50Ω straight-through clamp is 0.35mm, the thickness of the metal conductor line of the 50Ω straight-through clamp is 0.127mm, and the dielectric constant of the 50Ω straight-through clamp is 2.2; the length of the metal conductor line of the 150Ω straight-through clamp is 180mm, the width of the metal conductor line at both ends of the 150Ω straight-through clamp is 3.1mm, the middle width of the metal conductor line of the 150Ω straight-through clamp is 0.25mm, the thickness of the metal conductor line of the 150Ω straight-through clamp is 1.016mm, and the dielectric constant of the 150Ω straight-through clamp is 2.2.

[0017] Furthermore, the function expression of the metal conductor line profile of the 150Ω through fixture is:

[0018] y=Ae m(l-x)

[0019] Wherein x is the lateral length coordinate of the metal conductor line of the 150Ω through-fitting fixture; y is half the width of the metal conductor line of the 150Ω through-fitting fixture at the x coordinate; A is half the width of the metal conductor line of the corresponding 50Ω through-fitting fixture; e is a natural constant; m is a constant; l is half the length of the metal conductor line of the 150Ω through-fitting fixture.

[0020] Further, the open circuit calibration fixture includes a 50Ω open circuit fixture for current probe calibration and a 150Ω open circuit fixture for voltage probe calibration; wherein:

[0021] The 50Ω open-circuit fixture includes a first component and a second component, wherein the first component is a microstrip line structure and the second component is a ground terminal; the first component and the second component are in an open circuit state; the metal conductor line width of the first component matches the size of the probe tip to avoid additional distributed inductance and parasitic capacitance;

[0022] The 150Ω open-circuit fixture includes a third component and a fourth component. The third component is an impedance gradient microstrip line structure, and the fourth component is a ground terminal. The third component and the fourth component are in an open circuit state. The third component converts the 50Ω of the coaxial input terminal into an on-chip 150Ω characteristic impedance and matches it with the input impedance of the voltage probe to reduce reflections caused by impedance mismatch.

[0023] Furthermore, the length of the metal conductor line of the 50Ω open-circuit fixture is 45mm, the width of the metal conductor line of the 0Ω open-circuit fixture is 0.35mm, the thickness of the metal conductor line of the 0Ω open-circuit fixture is 0.127mm, and the dielectric constant of the 0Ω open-circuit fixture is 2.2; the length of the metal conductor line of the 150Ω open-circuit fixture is 90mm, the widths of the metal conductor line at both ends of the 150Ω open-circuit fixture are 3.1mm and 0.25mm respectively, the thickness of the metal conductor line of the 150Ω open-circuit fixture is 1.016mm, and the dielectric constant of the 150Ω open-circuit fixture is 2.2.

[0024] Furthermore, the specific method for performing the de-embedding calculation in step S5 includes the following sub-steps:

[0025] S5-1. The through calibration fixture is regarded as a cascade of two symmetrical microstrip line fixtures on the left and right. The two symmetrical microstrip line fixtures on the right are fixture A and fixture B. The scattering parameter matrix expression of fixture A is The scattering parameter matrix expression of fixture B is: S A11 , S A12 , S A21 and S A22 are the input reflection coefficient, reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture A respectively; S B11 , S B12 , S B21 and S B22 are the input reflection coefficient, reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture B respectively; S A21 =S A12 ; S B21 =S B12 ;

[0026] S5-2, calculating the S matrix of fixture A according to the input reflection coefficient of the open-circuit calibration fixture and the S matrix of the through-calibration fixture;

[0027] S5-3. Convert the S matrix of fixture A into a T matrix, denoted as T A ; Its expression is:

[0028]

[0029] Where |S A |=S A11 S A22 -S A12 S A21 ; T A is the T matrix of fixture A;

[0030] S5-4. Convert the overall S matrix into a T matrix, denoted as T M ; Its expression is:

[0031]

[0032] Where S M11 , S M21 and S M22 are the elements in the overall S matrix, representing the input reflection coefficient, forward transmission coefficient, and output reflection coefficient respectively; |S M |=S M11 S M22 -SM12 S M21 ;

[0033] S5-5, according to the formula:

[0034] [T DUT ]=[T A ] -1 ·[T M ]

[0035] Perform de-embedding calculation to obtain the T matrix T of the coupled probe under test DUT ;

[0036] S5-6, the coupled probe T matrix T to be tested DUT Convert to S matrix, denoted as S DUT , that is, the transfer function of the coupled probe under test is obtained.

[0037] Furthermore, the specific method of step S5-2 includes the following sub-steps:

[0038] S5-2-1. Equating the S matrix of the open-circuit calibration fixture with the S matrix of fixture A;

[0039] S5-2-2, performing inverse Fourier transform on the input reflection coefficient of the open-circuit calibration fixture to obtain the pulse response of the open-circuit calibration fixture;

[0040] S5-2-3, obtaining the time span from the start to the maximum pulse peak from the pulse response of the open-circuit calibration fixture, and using the time span as the width of the time domain gate to perform time domain wave interception on the input reflection coefficient of the open-circuit calibration fixture, to obtain the time domain characteristics of the open-circuit calibration fixture;

[0041] S5-2-4. Convert the time domain characteristics of the open-circuit calibration fixture into frequency form through Fourier transform to obtain the input reflection coefficient S of fixture A. A11 ;

[0042] S5-2-5, according to the formula:

[0043]

[0044] The reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture A are obtained, and then the S matrix of fixture A is obtained; where S A11-opne is the input reflection coefficient of the open-circuit calibration fixture; S T11 , S T21 and S T22 are the elements in the S matrix of the through calibration fixture, representing the input reflection coefficient, forward transmission coefficient, and output reflection coefficient of the through calibration fixture respectively.

[0045] The beneficial effects of the present invention are:

[0046] 1. This method can quickly and effectively embed the influence of the calibration fixture and obtain accurate and repeatable measurement results, thereby accurately and quickly calibrating the direct-coupled probe.

[0047] 2. This method designs a corresponding impedance transformation calibration fixture based on the special structure of DCP, which is different from the microwave probe, and can improve the calibration capability of DCP to 20GHz.

[0048] 3. This method can effectively suppress the jump and ripple of the high-frequency part of the measurement result caused by the impedance mismatch between the measurement system and the probe, improve the accuracy and repeatability of the probe calibration, and can effectively make up for the problems of insufficient current probe calibration frequency range and lack of voltage probe calibration method in standard IEC 61967-4. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a schematic diagram of the process of this method;

[0050] Figure 2 It is a schematic diagram of the structure of a 50Ω through calibration fixture and a 50Ω open-circuit calibration fixture;

[0051] Figure 3 It is a schematic diagram of the structure of a 150Ω through calibration fixture and a 150Ω open-circuit calibration fixture;

[0052] Figure 4 It is a schematic diagram of the connection between the open-circuit calibration fixture, the vector network analyzer and the coupled probe under test;

[0053] Figure 5 This is the simulation result of 150Ω straight-through fixture (impedance transformation fixture);

[0054] Figure 6 This is the simulation result of 50Ω through fixture;

[0055] Figure 7 The schematic diagram of the experimental connection for the calibration according to the standard IEC 61967-4 in the embodiment;

[0056] Figure 8 is the transmission coefficient of DCP calculated by principle circuit simulation;

[0057] Fig. 9 is the DCP transmission coefficient measured using the standard method;

[0058] Fig.10 This is the amplitude comparison result of 150Ω voltage probe S21;

[0059] Fig.11 This is the phase comparison result of 150Ω voltage probe S21;

[0060] Fig.12 This is the comparison result of the amplitude of 1Ω current probe S21;

[0061] Fig.13 This is the phase comparison result of 1Ω current probe S21;

[0062] Fig.14 The effect of 150Ω through-fixture grounding on the measurement results;

[0063] Fig.15 The effect of grounding the current probe with 1Ω impedance on the measurement results;

[0064] Fig.16 Comparison of the measurement results of 50Ω and 150Ω fixtures;

[0065] Fig.17 The effect of time window selection on measurement; (a) is the TDR waveform of the fixture; (b) is the standard deviation of the measurement results with different time window lengths. DETAILED DESCRIPTION

[0066] The specific implementation modes of the present invention are described below to facilitate those skilled in the art to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0067] like Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown, the calibration method of the ultra-wideband conducted emission direct coupling probe includes the following steps:

[0068] S1. Perform a dual-port calibration on a vector network analyzer, and calibrate the reference end face to the port of the coaxial cable to be measured; wherein the vector network analyzer is a dual-port vector network analyzer;

[0069] S2. Connect a straight-through calibration fixture between two ports of the vector network analyzer and measure the S matrix of the straight-through calibration fixture;

[0070] S3, connecting an open-circuit calibration fixture only to the first port of the vector network analyzer, and measuring the input reflection coefficient of the open-circuit calibration fixture;

[0071] S4, maintaining the connection between the open-circuit calibration fixture and the first port of the vector network analyzer, connecting the ground terminal of the measured coupling probe to the ground terminal of the open-circuit calibration fixture, connecting the input terminal of the measured coupling probe to the output terminal of the open-circuit calibration fixture, connecting the output terminal of the measured coupling probe to the second port of the vector network analyzer, and measuring to obtain an overall S matrix formed by the open-circuit calibration fixture and the measured coupling probe;

[0072] S5, performing de-embedding calculation based on the S matrix of the through calibration fixture, the input reflection coefficient of the open-circuit calibration fixture and the overall S matrix to obtain the transfer function of the coupled probe under test;

[0073] S6. Calibrate the coupled probe under test based on the transfer function of the coupled probe under test.

[0074] In this embodiment, since the probe has a metal tip at the input end and a metal plane grounded at the bottom instead of a coaxial structure, it cannot be directly connected to the measuring device, and an appropriate physical structure is required to convert its input end into a coaxial structure. Also, due to the nonlinear characteristics of the internal components of the probe, it does not have the reciprocity principle of a general microwave passive network, so it is not possible to use a "back-to-back" method to contact the tips of the two probes for reciprocity measurement. Therefore, this embodiment designs a calibration fixture based on a microstrip structure:

[0075] The through-calibration fixture includes a 50Ω through-calibration fixture for current probe calibration and a 150Ω through-calibration fixture for voltage probe calibration; wherein:

[0076] The 50Ω through fixture is a microstrip line structure, and the metal conductor line width matches the size of the probe tip to avoid additional distributed inductance and parasitic capacitance;

[0077] The 150Ω through fixture is an impedance gradient microstrip line structure, which converts the 50Ω at the coaxial input end into the 150Ω characteristic impedance on the chip and matches it with the input impedance of the voltage probe to reduce reflections caused by impedance mismatch.

[0078] The length of the metal conductor line of the 50Ω straight-through clamp is 90mm, the width of the metal conductor line of the 50Ω straight-through clamp is 0.35mm, the thickness of the metal conductor line of the 50Ω straight-through clamp is 0.127mm, and the dielectric constant of the 50Ω straight-through clamp is 2.2; the length of the metal conductor line of the 150Ω straight-through clamp is 180mm, the width of the metal conductor line at both ends of the 150Ω straight-through clamp is 3.1mm, the middle width of the metal conductor line of the 150Ω straight-through clamp is 0.25mm, the thickness of the metal conductor line of the 150Ω straight-through clamp is 1.016mm, and the dielectric constant of the 150Ω straight-through clamp is 2.2.

[0079] like Figure 5 and Figure 6As shown, the S parameter characteristics of the two structures can meet the basic test requirements at frequencies above 1 GHz, where the S11 performance of the two structures is relatively close, while the S21 parameter of the linear structure is relatively poor. When the insertion loss / return loss of the measurement fixture is large, the error introduced by the de-embedding calculation will be amplified. When there is a difference in workmanship between the calibrated fixture and the actual measurement fixture, if the insertion loss / return loss of the calibration fixture is large, the error introduced by the de-embedding calculation will be amplified. Therefore, under the premise of ensuring the consistency of the workmanship of the calibration fixture, this embodiment selects an exponential impedance transformation structure because its S parameter performance is relatively better.

[0080] Therefore, the function expression of the metal conductor line profile of the 150Ω through fixture is:

[0081] y=Ae m(l-x)

[0082] Wherein x is the lateral length coordinate of the metal conductor line of the 150Ω through-fitting fixture; y is half the width of the metal conductor line of the 150Ω through-fitting fixture at the x coordinate; A is half the width of the metal conductor line of the corresponding 50Ω through-fitting fixture; e is a natural constant; m is a constant; l is half the length of the metal conductor line of the 150Ω through-fitting fixture.

[0083] The open circuit calibration fixture includes a 50Ω open circuit fixture for current probe calibration and a 150Ω open circuit fixture for voltage probe calibration; wherein:

[0084] The 50Ω open-circuit fixture includes a first component and a second component, wherein the first component is a microstrip line structure and the second component is a ground terminal; the first component and the second component are in an open circuit state; the metal conductor line width of the first component matches the size of the probe tip to avoid additional distributed inductance and parasitic capacitance;

[0085] The 150Ω open-circuit fixture includes a third component and a fourth component. The third component is an impedance gradient microstrip line structure, and the fourth component is a ground terminal. The third component and the fourth component are in an open circuit state. The third component converts the 50Ω of the coaxial input terminal into an on-chip 150Ω characteristic impedance and matches it with the input impedance of the voltage probe to reduce reflections caused by impedance mismatch.

[0086] The length of the metal conductor line of the 50Ω open-circuit fixture is 45mm, the width of the metal conductor line of the 0Ω open-circuit fixture is 0.35mm, the thickness of the metal conductor line of the 0Ω open-circuit fixture is 0.127mm, and the dielectric constant of the 0Ω open-circuit fixture is 2.2; the length of the metal conductor line of the 150Ω open-circuit fixture is 90mm, the widths of the metal conductor line at both ends of the 150Ω open-circuit fixture are 3.1mm and 0.25mm respectively, the thickness of the metal conductor line of the 150Ω open-circuit fixture is 1.016mm, and the dielectric constant of the 150Ω open-circuit fixture is 2.2.

[0087] In the specific implementation process, there are many ways to realize the structure of microstrip line impedance transformation, including step type, linear type, exponential type, etc. Different impedance transformation structures have different effects on the transmission characteristics of the calibration piece, especially in the high-frequency part. Since the step type has a discontinuous change point of impedance in the time domain, a large standing wave will be generated at a high frequency. This embodiment simulates and compares the scattering parameters of linear and exponential impedance transformation microstrip lines.

[0088] Unlike the existing methods, this method only needs one fixture instead of a pair of fixtures in the 2X-Thru method during the measurement of the conducted emission probe, so the electrical length and transmission characteristics of the single fixture need to be accurately obtained. And due to the particularity and nonlinear characteristics of the probe structure, it cannot be assumed to be reciprocal. This increases the number of variables that need to be solved when using the 2X-Thru algorithm, and this method needs to introduce additional equations for solution. This method uses a TDR-based 1X / 2X-THRU de-embedding time-frequency domain algorithm, and adds an open-circuit calibration component on the basis of the 2X-Thru through calibration component, so as to accurately solve the transmission characteristics of the fixture. Specifically, the specific method for performing de-embedding calculations in step S5 includes the following sub-steps:

[0089] S5-1. The through calibration fixture is regarded as a cascade of two symmetrical microstrip line fixtures on the left and right. The two symmetrical microstrip line fixtures on the right are fixture A and fixture B. The scattering parameter matrix expression of fixture A is The scattering parameter matrix expression of fixture B is: S A11 , S A12 , S A21 and S A22 are the input reflection coefficient, reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture A respectively; S B11 , S B12 , S B22 and S B22 are the input reflection coefficient, reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture B respectively; S A21 =S A12 ; S B21 =S B12 ; The scattering parameter matrix of the open-circuit calibration fixture in the case of a well-fitted connection should be consistent with that of fixture A; when the open-circuit calibration fixture is not connected and remains in an open-circuit state, its input reflection coefficient will change. The input reflection coefficient measured by the vector network analyzer (VNA) is recorded as S A11-opne ;

[0090] S5-2, calculating the S matrix of fixture A according to the input reflection coefficient of the open-circuit calibration fixture and the S matrix of the through-calibration fixture;

[0091] S5-3. Convert the S matrix of fixture A into a T matrix, denoted as T A ; Its expression is:

[0092]

[0093] Where |S A |=S A11 S A22 -S A12 S A21 ; T A is the T matrix of fixture A;

[0094] S5-4. Convert the overall S matrix into a T matrix, denoted as T M ; Its expression is:

[0095]

[0096] Where S M11 , S M21 and S M22 are the elements in the overall S matrix, representing the input reflection coefficient, forward transmission coefficient, and output reflection coefficient respectively; |S M |=S M11 S M22 -S M12 S M21 ;

[0097] S5-5, according to the formula:

[0098] [T DUT ]=[T A ] -1 ·[T M ]

[0099] Perform de-embedding calculation to obtain the T matrix T of the coupled probe under test DUT ;

[0100] S5-6, the coupled probe T matrix T to be tested DUT Convert to S matrix, denoted as S DUT , that is, the transfer function of the coupled probe under test is obtained.

[0101] The specific method of step S5-2 includes the following sub-steps:

[0102] S5-2-1. Equating the S matrix of the open-circuit calibration fixture with the S matrix of fixture A;

[0103] S5-2-2, performing inverse Fourier transform on the input reflection coefficient of the open-circuit calibration fixture to obtain the pulse response of the open-circuit calibration fixture;

[0104] S5-2-3, obtaining the time span from the start to the maximum pulse peak from the pulse response of the open-circuit calibration fixture, and using the time span as the width of the time domain gate to perform time domain wave interception on the input reflection coefficient of the open-circuit calibration fixture, to obtain the time domain characteristics of the open-circuit calibration fixture;

[0105] S5-2-4. Convert the time domain characteristics of the open-circuit calibration fixture into frequency form through Fourier transform to obtain the input reflection coefficient S of fixture A. A11 ;

[0106] S5-2-5, according to the formula:

[0107]

[0108] The reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture A are obtained, and then the S matrix of fixture A is obtained; where S A11-opne is the input reflection coefficient of the open-circuit calibration fixture; S T11 , S T21 and S T22 are the elements in the S matrix of the through calibration fixture, representing the input reflection coefficient, forward transmission coefficient, and output reflection coefficient of the through calibration fixture respectively.

[0109] In the specific implementation process, if the VNA used for measurement does not have a time domain option, it is necessary to perform a Fourier transform on the data collected by the VNA in the time-frequency domain. Since the data transformation process belongs to discrete Fourier transform (DFT), when performing VNA measurement, the starting frequency of the VNA and the measurement frequency interval need to be set to be the same, so as to match the requirements of the DFT process on the computer for the data form. After the fixture time domain data is intercepted, it is necessary to perform a zero padding operation on the intercepted data to improve the accuracy of the Fourier transform calculation. When intercepting data, a suitable time window should be selected. If a rectangular window is directly used for interception, a truncation effect will occur after the FFT is converted to the frequency domain. This embodiment can use weighted window functions such as Gaussian window and Hanning window.

[0110] In one embodiment of the present invention, a calibration method for a 1Ω current probe based on a PCB structure is provided in Appendix A of the standard IEC 61967-4, and the calibration frequency covers 150kHz-1GHz, but a calibration method for a 150Ω voltage probe is not provided. The experimental connection principle of the calibration is as follows: Figure 7 As shown in the figure, a vector network analyzer (or a swept signal source plus a signal receiver) is used within the calibration frequency band to connect a 50Ω standard load, and a current probe is connected across the load. The RF source is switched to the input end of the probe to obtain the insertion loss of the current probe. This measurement result is the insertion loss of the RF current probe, indicating its sensitivity.

[0111] According to the circuit structure of 1Ω / 150Ω probe, the theoretical transmission coefficient of current probe and voltage probe can be obtained as follows: Figure 8 When the input impedance of a 1Ω current probe is 50Ω, its theoretical transmission coefficient is 34dB-36dB. When the input impedance of a 150Ω current probe is 50Ω, its theoretical transmission coefficient is close to -11.7dB; when the input impedance is 150Ω, its theoretical transmission coefficient is around -10.6dB.

[0112] As a comparison with the theoretical calculated value, this embodiment adopts the method described in the standard to measure the transmission coefficient (insertion loss) of the commercial probe using a network analyzer. The results of the calibration measurement for the existing commercial current probe (Langer P750 / P603-1) are as follows: Fig. 9 The operating frequency range of existing commercial probes is 150kHz-3GHz, which exceeds the operating range of the method specified in the standard. Therefore, it can be seen that the data obtained by the measurement method in the standard has produced obvious oscillations above the frequency of 1G, which is quite different from the calculated theoretical transmission coefficient of the probe.

[0113] Therefore, the current calibration method of 1Ω / 150Ω conductive probe has the following deficiencies that need to be improved:

[0114] (a) The standard does not provide a calibration method for a 150Ω voltage conduction probe;

[0115] (b) The calibration method given in the standard cannot cover the operating frequency of existing commercial probes;

[0116] (c) The influence introduced by the PCB / coaxial calibration fixture cannot be evaluated and removed;

[0117] (d) The repeatability and stability of the measurement cannot be guaranteed.

[0118] This embodiment uses a vector network analyzer Keysight E5071C to measure the commercial conducted emission direct coupling probe Langer P750 / P603-1, which has a nominal operating frequency range of 150kHz-3GHz. This embodiment compares the measurement results of the standard method and this method. The voltage probe uses a 150Ω straight-through fixture, and the current probe uses a 50Ω straight-through fixture. Fig.10 , Fig.11 , Fig.12 and Fig.13As shown in the figure, after using this method, the reflection and ripple caused by mismatch can be significantly suppressed, and the probe attenuation value is also corrected, and the result is closer to the theoretical calculated value. It can be seen that the current probe is more sensitive than the voltage probe, and factors such as mismatch and distributed inductance will bring drastic changes to the measurement results, so it is particularly important to use a suitable calibration fixture.

[0119] Due to the probe structure and experimental arrangement, the probe may not be in sufficient contact with the ground plane of the calibration fixture during the test. This embodiment uses a probe constructed with a self-made equivalent circuit. During the measurement process, the area of ​​contact between the probe ground plane and the ground end of the calibration fixture is intentionally reduced to simulate the situation of insufficient probe grounding, so as to evaluate the impact of grounding quality on the measurement results. Fig.14 and Fig.15 As shown in the figure, whether the probe is adequately grounded has a great impact on the measurement results. The 1Ω impedance current probe has a small input impedance and is affected by parasitic capacitance and other factors. There is a characteristic deviation at a frequency of about 2GHz. After multiple experiments, the deviation is stably reproduced. Through analysis, it may be due to the resonance point caused by the size of the probe itself. The 150Ω probe has a large input impedance, and the impact on the low frequency part is smaller than that of the current probe, but it will produce large oscillations in the high frequency part.

[0120] This example also compares the difference in measurement results when the voltage probe is measured using a 150Ω impedance conversion calibration fixture and a 50Ω impedance calibration fixture:

[0121] A commercial conducted emission voltage probe (the product nominal range is 150kHz-3GHz, but in order to verify the difference in measurement results under a wider frequency band, this embodiment widens the measurement frequency range) is used for comparative measurement. Fig.16 As shown in the figure, there is a large system deviation between the two measurement methods below 1GHz, with the maximum difference being 1.4dB. The reason for this is as mentioned above. At 1GHz, the calibration fixture has a large insertion loss, which amplifies the measurement uncertainty introduced. Above 2GHz, the overall trend of the measurement results of the two fixtures is relatively consistent, but the measurement result curve of the 150Ω impedance transformation fixture is smoother. This method is more conducive to suppressing the S parameter oscillation caused by the mismatch reflection at the probe tip, and the effect is more significant in the high-frequency part.

[0122] Since this method needs to obtain the S parameters of the calibration fixture by time domain interception according to the time point when the peak of the through / open calibration fixture pulse response appears, when judging the time window range to be intercepted, the interception time length may be different due to the algorithm or human selection, which will eventually lead to deviation in the measurement result. In order to judge the impact of time window length interception on the measurement result, this embodiment selects several time points near the physical center of the calibration fixture for interception, and the comparison results are as follows: Fig.17 shown.

[0123] pass Fig.17 The TDR response in (a) shows that the time domain impedance of the through fixture changes from 50Ω to 150Ω and then back to 150Ω. The highest point of the TDR impedance change is the center of the calibration fixture, which is also the time window length (1.3ns) that needs to be set in this method. This embodiment randomly selects 9 time points within the range of ±0.1ns near the center of the calibration fixture as the time window length for comparison to perform de-embedding calculations. By calculating different measurement data, the following is obtained: Fig.17 The standard deviation result shown in (b) has a maximum standard deviation of less than 0.25 dB. It can be seen that intercepting the time window near the electrical length of the calibration fixture will cause a certain deviation in the measurement results, but the uncertainty can be controlled within an acceptable range.

[0124] In summary, the present invention proposes a fixture de-embedding method based on the combination of TDR and frequency domain, and designs and develops an impedance adaptation calibration fixture. Through simulation and measurement verification of the fixture, this method can achieve the measurement of the transmission characteristics of the probe covering 20GHz, which better meets the needs of the existing integrated circuit field for conducted emission measurement.

Claims

1. A calibration method for an ultra-wideband conducted emission direct coupling probe, characterized in that: The following steps are involved: S1. Perform a dual-port calibration on a vector network analyzer, and calibrate the reference end face to the port of the coaxial cable to be measured; wherein the vector network analyzer is a dual-port vector network analyzer; S2. Connect a straight-through calibration fixture between two ports of the vector network analyzer and measure the S matrix of the straight-through calibration fixture; S3, connecting an open-circuit calibration fixture only to the first port of the vector network analyzer, and measuring the input reflection coefficient of the open-circuit calibration fixture; S4, maintaining the connection between the open-circuit calibration fixture and the first port of the vector network analyzer, connecting the ground terminal of the measured coupling probe to the ground terminal of the open-circuit calibration fixture, connecting the input terminal of the measured coupling probe to the output terminal of the open-circuit calibration fixture, connecting the output terminal of the measured coupling probe to the second port of the vector network analyzer, and measuring to obtain an overall S matrix formed by the open-circuit calibration fixture and the measured coupling probe; S5, performing de-embedding calculation based on the S matrix of the through calibration fixture, the input reflection coefficient of the open-circuit calibration fixture and the overall S matrix to obtain the transfer function of the coupled probe under test; S6. Calibrate the coupled probe under test based on the transfer function of the coupled probe under test.

2. The calibration method of the ultra-wideband conducted emission direct coupling probe according to claim 1, characterized in that: The through-calibration fixture includes a 50Ω through-calibration fixture for current probe calibration and a 150Ω through-calibration fixture for voltage probe calibration; wherein: The 50Ω through fixture is a microstrip line structure, and the metal conductor line width matches the size of the probe tip to avoid additional distributed inductance and parasitic capacitance; The 150Ω through fixture is an impedance gradient microstrip line structure, which converts the 50Ω of the coaxial input end into the 150Ω characteristic impedance on the chip and matches it with the input impedance of the voltage probe to reduce reflections caused by impedance mismatch.

3. The calibration method of the ultra-wideband conducted emission direct coupling probe according to claim 2, characterized in that: The length of the metal conductor line of the 50Ω straight-through clamp is 90mm, the width of the metal conductor line of the 50Ω straight-through clamp is 0.35mm, the thickness of the metal conductor line of the 50Ω straight-through clamp is 0.127mm, and the dielectric constant of the 50Ω straight-through clamp is 2.2; the length of the metal conductor line of the 150Ω straight-through clamp is 180mm, the width of the metal conductor line at both ends of the 150Ω straight-through clamp is 3.1mm, the middle width of the metal conductor line of the 150Ω straight-through clamp is 0.25mm, the thickness of the metal conductor line of the 150Ω straight-through clamp is 1.016mm, and the dielectric constant of the 150Ω straight-through clamp is 2.

2.

4. The calibration method of the ultra-wideband conducted emission direct coupling probe according to claim 2, characterized in that: The function expression of the metal conductor line profile of the 150Ω through fixture is: in is the horizontal length coordinate of the metal conductor line of the 150Ω through fixture; The metal conductor line of the 150Ω straight-through fixture is half width at coordinates; is half the width of the metal conductor line corresponding to the 50Ω through fixture; e is a natural constant; is a constant; Half the length of the metal conductor line of the 150Ω thru fixture.

5. The calibration method of the ultra-wideband conducted emission direct coupling probe according to claim 2, characterized in that: The open circuit calibration fixture includes a 50Ω open circuit fixture for current probe calibration and a 150Ω open circuit fixture for voltage probe calibration; wherein: The 50Ω open-circuit fixture includes a first component and a second component, wherein the first component is a microstrip line structure and the second component is a ground terminal; the first component and the second component are in an open circuit state; the metal conductor line width of the first component matches the size of the probe tip to avoid additional distributed inductance and parasitic capacitance; The 150Ω open-circuit fixture includes a third component and a fourth component. The third component is an impedance gradient microstrip line structure, and the fourth component is a ground terminal. The third component and the fourth component are in an open circuit state. The third component converts the 50Ω of the coaxial input terminal into an on-chip 150Ω characteristic impedance and matches it with the input impedance of the voltage probe to reduce reflections caused by impedance mismatch.

6. The calibration method of the ultra-wideband conducted emission direct coupling probe according to claim 5, characterized in that: The length of the metal conductor line of the 50Ω open-circuit fixture is 45mm, the width of the metal conductor line of the 50Ω open-circuit fixture is 0.35mm, the thickness of the metal conductor line of the 50Ω open-circuit fixture is 0.127mm, and the dielectric constant of the 50Ω open-circuit fixture is 2.2; the length of the metal conductor line of the 150Ω open-circuit fixture is 90mm, the widths of the metal conductor line at both ends of the 150Ω open-circuit fixture are 3.1mm and 0.25mm respectively, the thickness of the metal conductor line of the 150Ω open-circuit fixture is 1.016mm, and the dielectric constant of the 150Ω open-circuit fixture is 2.

2.

7. The calibration method of the ultra-wideband conducted emission direct coupling probe according to claim 1, characterized in that: The specific method for performing the de-embedding calculation in step S5 includes the following sub-steps: S5-1. The straight-through calibration fixture is regarded as a cascade of two symmetrical microstrip line fixtures on the left and right. The two symmetrical microstrip line fixtures on the left and right are fixture A and fixture B. The scattering parameter matrix expression of fixture A is ; The scattering parameter matrix expression of fixture B is: ; , , and are the input reflection coefficient, reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture A respectively; , , and are the input reflection coefficient, reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture B respectively; ; ; S5-2, calculating the S matrix of fixture A according to the input reflection coefficient of the open-circuit calibration fixture and the S matrix of the through-calibration fixture; S5-3. Convert the S matrix of fixture A into a T matrix, recorded as ; Its expression is: in ; is the T matrix of fixture A; S5-4, convert the overall S matrix into a T matrix, recorded as ; Its expression is: in , and are the elements in the overall S matrix, representing the input reflection coefficient, forward transmission coefficient, and output reflection coefficient respectively; ; S5-5, according to the formula: Perform de-embedding calculation to obtain the T matrix of the coupled probe under test ; S5-6. T matrix of the coupled probe to be tested Convert to S matrix, denoted as , that is, the transfer function of the coupled probe under test is obtained.

8. The calibration method of the ultra-wideband conducted emission direct coupling probe according to claim 7, characterized in that: The specific method of step S5-2 includes the following sub-steps: S5-2-1. Equating the S matrix of the open-circuit calibration fixture with the S matrix of fixture A; S5-2-2, performing inverse Fourier transform on the input reflection coefficient of the open-circuit calibration fixture to obtain the pulse response of the open-circuit calibration fixture; S5-2-3, obtaining the time span from the start to the maximum pulse peak from the pulse response of the open-circuit calibration fixture, and using the time span as the width of the time domain gate to perform time domain wave interception on the input reflection coefficient of the open-circuit calibration fixture, to obtain the time domain characteristics of the open-circuit calibration fixture; S5-2-4. Convert the time domain characteristics of the open-circuit calibration fixture into frequency form through Fourier transform to obtain the input reflection coefficient of fixture A. ; S5-2-5, according to the formula: The reverse transmission coefficient, forward transmission coefficient and output reflection coefficient of fixture A are obtained, and then the S matrix of fixture A is obtained; is the input reflection coefficient of the open circuit calibration fixture; , and are the elements in the S matrix of the through calibration fixture, representing the input reflection coefficient, forward transmission coefficient, and output reflection coefficient of the through calibration fixture respectively.

Images