Passive intermodulation prediction method based on normal distribution wideband signal

Abstract

The present application relates to the technical field of radio frequency passive intermodulation test, and relates to a passive intermodulation prediction method based on normal distribution wideband signal, which comprises the following steps: step 1: through double-tone intermodulation test, the third, fifth and seventh order passive intermodulation of the connector to be tested is tested; step 2: the contact resistance of the connector to be tested is tested; step 3: the nonlinear polynomial model parameters are calculated according to the test results; step 4: a wideband signal mathematical model based on normal distribution is established; step 5: a nonlinear polynomial mathematical model of the radio frequency connector is established with the wideband signal as an excitation source; the power of the wideband signal passive intermodulation is predicted through the nonlinear polynomial mathematical model, and the results are analyzed. The present application can preferably predict passive intermodulation.

Description

Technical Field

[0001] The present invention relates to the technical field of radio frequency passive intermodulation testing, in particular to a passive intermodulation prediction method based on normally distributed broadband signals. Background Art

[0002] When two or more frequency signals pass through a passive device, the nonlinear characteristics of the passive device cause signal mixing, generating frequency signals other than the input signal frequency. This interference phenomenon is called passive intermodulation. This phenomenon is common in various communication systems, especially in the design of high-power multi-channel satellite communication systems. The problem of passive intermodulation is very important, and the passive intermodulation level of the entire system needs to be strictly controlled and evaluated. Traditional methods for testing passive intermodulation use two continuous waves of equal power as signal sources for testing. This can effectively measure the nonlinear characteristics of passive components. However, in modern wireless communications, broadband signals are widely used, and traditional test methods cannot accurately predict the nonlinear distortion of components under broadband signal excitation.

[0003] The traditional method for testing passive intermodulation (two-tone intermodulation test) uses two continuous waves (CW) of equal power as the signal source for testing. However, in modern wireless communications, broadband signals are widely used. Traditional test methods cannot use broadband signals as signal sources for testing, cannot effectively reflect the passive intermodulation of broadband signals, and cannot predict the nonlinear distortion of passive components under broadband signal excitation. Existing passive intermodulation test systems for broadband signals have complex structures, are difficult to assemble and disassemble, and are not convenient for long-term use. Summary of the Invention

[0004] The present invention provides a passive intermodulation prediction method based on normally distributed broadband signals, which can solve the problems that traditional methods for testing passive intermodulation cannot use broadband signals as signal sources for testing, cannot effectively analyze the behavioral characteristics of broadband signal passive intermodulation, and the existing passive intermodulation test systems for broadband signals are complex in structure and not convenient for long-term use.

[0005] According to the present invention, the passive intermodulation prediction method based on the normally distributed broadband signal comprises the following steps:

[0006] Step 1: Use the two-tone intermodulation test to test the third, fifth, and seventh order passive intermodulation of the connector under test;

[0007] Step 2: Test the contact resistance of the connector to be tested;

[0008] Step 3: Calculate the nonlinear polynomial model parameters based on the test results;

[0009] Step 4: Establish a mathematical model of broadband signals based on normal distribution;

[0010] Step 5: Establish a nonlinear polynomial mathematical model of the RF connector using the broadband signal as the excitation source; predict the power of the broadband signal passive intermodulation through the nonlinear polynomial mathematical model and analyze the results.

[0011] As preferably, in step 1, JCIMA-900-P passive intermodulation analyzer is used to test, when performing passive intermodulation test, it is necessary for the self-intermodulation of the passive intermodulation instrument to be lower than -120dBm, and each measurement is carried out in a room where the air-conditioning temperature is 20 degrees Celsius, to ensure indoor temperature consistency, a torque wrench is used to reduce the influence of connector tightness, and the error during intermodulation test is reduced by the method for averaging by multiple measurements, and the instrument warm-up time is at least 3 minutes, and the signal transmission frequency band is 930-960MHz, and the intermodulation receiving frequency band is 885-915MHz.

[0012] Preferably, in step 1, during the seventh-order passive intermodulation test, the input signal frequencies are set to 933 MHz and 949 MHz, and the magnitude of the seventh-order passive intermodulation at 885 MHz is measured.

[0013] Preferably, in step 2, if the connector to be tested is an N-type connector, the contact resistance model is:

[0014] The overall resistance R1 includes the contact resistance R2 of the two copper conductors and the connector, the connector contact resistance R, and the resistance R3 of the device being measured.

[0015] R1=R+2R2+R3(1)

[0016] Use a JK2511 DC resistance tester to measure the overall resistance R1. Then connect the two copper conductors and measure their contact resistance. The measurement result is considered to be the contact resistance R2 between the copper conductor and the connector. Use a vernier caliper to measure the effective length L and the cross-sectional diameter d of the N-type connector. According to the resistance determination formula (2), the N-type connection resistance R3 is calculated:

[0017]

[0018] Where ρ is the resistivity of copper;

[0019] According to formula (1), the contact resistance of the coaxial connector is:

[0020] R=R1-2R2-R3(3).

[0021] Preferably, in step 3, a memoryless nonlinear polynomial model is selected to describe the current-voltage IV characteristic curve of the connector:

[0022]

[0023] Where i is the current through the connector, v is the voltage across the connector, and a k are the nonlinear polynomial model coefficients of the connector;

[0024] For a two-tone CW signal, the input voltage is expressed as:

[0025] v=V[cos(2πf1t)+cos(2πf2t)](5)

[0026] Where V is the carrier amplitude, f1 and f2 are the carrier frequencies;

[0027] The current-voltage IV characteristic curve retained to the 7th order is:

[0028] i=a1v+a3v 3 +a5v 5 +a7v 7 (6)

[0029] The nonlinear polynomial model coefficients of the coaxial connector are measured through a two-tone test experiment.

[0030] Substituting (5) into (6), we get the expression for the current, extracting the term with a frequency of 2f1-f2, and calculating the current of the third-order intermodulation product. Similarly, we can obtain the current of the fifth-order and seventh-order products:

[0031]

[0032]

[0033]

[0034] In the nonlinear model of the connector, a1 is the contact conductance, that is:

[0035]

[0036] The relationship between current, resistance and power P is:

[0037] P=i 2 R load (11)

[0038] R load is the load resistance;

[0039] Using the power of the third-order, fifth-order, and seventh-order intermodulation signals tested in step 2, the values ​​of a7, a5, and a3 are obtained according to equations (7)(8)(9)(10):

[0040]

[0041]

[0042]

[0043]

[0044] As a preference, in step 4, the broadband signal is modeled, assuming that the sidebands of the two broadband signals do not overlap, and each broadband signal is approximately equivalent to a CW signal with equal frequency intervals. The farther the CW signal is from the center frequency, the smaller the amplitude. According to the characteristics of the broadband signal, the amplitude of the broadband signal varies according to a normal distribution. Assume that a broadband signal is simulated by n CW signals, and the amplitudes of the n signals are V1, V2, ..., V n , obeys the normal distribution, that is:

[0045] V~N(μ,σ 2 )(16) where μ is the average amplitude of the broadband signal, σ 2 is the variance of the amplitude, N is the normal distribution;

[0046] The power P of each CW signal in the broadband signal based on normal distribution i According to power, voltage, load resistance R load The relationship is calculated as follows:

[0047]

[0048] The signal source is set to two broadband signals with the same power and center frequencies f1 and f2. The input voltage is expressed as:

[0049]

[0050]

[0051] v=v1+v2(20)

[0052] Where v1 represents the voltage of the first broadband signal, V 1,i 、f 1,i represents the voltage amplitude and frequency of the first broadband signal, the i-th CW signal, v2 represents the voltage of the second broadband signal, V 2,j 、f 2,j Represents the voltage amplitude and frequency of the j-th CW signal of the second broadband signal.

[0053] As a preference, in step 5, the broadband signal model (20) is substituted into (6) to obtain the expression of broadband passive intermodulation current, and the frequency 2f is extracted. 1,i -f 2,j , calculate the current of the third-order intermodulation product. Similarly, the current of the fifth-order and seventh-order products is obtained. According to the relationship between current, resistance and power, the corresponding passive intermodulation power is calculated.

[0054]

[0055]

[0056]

[0057] The present invention uses the test results of a two-tone intermodulation test to establish a passive nonlinear transmission mathematical model of a radio frequency connector, and uses an equivalent broadband signal based on a normal distribution as an excitation source of the nonlinear transmission model to perform simulation, and analyzes the behavioral characteristics of the broadband passive intermodulation according to the simulation results. The present invention can effectively analyze the behavioral characteristics of the passive intermodulation of broadband signals, and the passive intermodulation test system for broadband signals has a simple structure and can be used for a long time. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 Flowchart of a passive intermodulation prediction method based on a normally distributed broadband signal in an embodiment;

[0059] Figure 2 It is a two-tone intermodulation test spectrum diagram in the embodiment;

[0060] Figure 3 A schematic diagram of a customized copper conductor model in the embodiment;

[0061] Figure 4 Schematic diagram of the contact resistance model in the embodiment;

[0062] Figure 5 Schematic diagram of third-order passive intermodulation of an N-type connector in an embodiment;

[0063] Figure 6 Schematic diagram of the difference between signal power and third-order intermodulation power in an embodiment;

[0064] Figure 7 Schematic diagram of the circuit model and related parameters in the embodiment;

[0065] Figure 8 Schematic diagram of broadband signal passive intermodulation simulation in an embodiment;

[0066] Figure 9 Schematic diagram of broadband signal passive intermodulation simulation in an embodiment;

[0067] Figure 10 Schematic diagram of broadband signal power with a signal bandwidth of B=20 MHz and a frequency interval of d=2 MHz in an embodiment;

[0068] FIG11( a ) is a spectrum diagram of passive intermodulation products with a signal bandwidth B=20 MHz and a frequency interval d=2 MHz in an embodiment;

[0069] FIG11( b ) is a spectrum diagram of passive intermodulation products with a signal bandwidth B=20 MHz and a frequency interval d=1 MHz in an embodiment;

[0070] FIG11( c ) is a spectrum diagram of passive intermodulation products with a signal bandwidth of B=40 MHz and a frequency interval of d=2 MHz in the embodiment;

[0071] FIG11( d ) is a spectrum diagram of passive intermodulation products with a signal bandwidth of B=40 MHz and a frequency interval of d=1 MHz in an embodiment;

[0072] Figure 12 Schematic diagram of third-order passive intermodulation of each connector in the embodiment;

[0073] Figure 13 Schematic diagram of the difference between signal power and third-order intermodulation power in an embodiment. DETAILED DESCRIPTION

[0074] In order to further understand the content of the present invention, the present invention is described in detail with reference to the accompanying drawings and embodiments. It should be understood that the embodiments are merely for explaining the present invention and are not intended to limit the present invention.

[0075] Example

[0076] like Figure 1 As shown, this embodiment provides a passive intermodulation prediction method based on a normally distributed broadband signal, which includes the following steps:

[0077] Step 1: Use the two-tone intermodulation test to test the third, fifth, and seventh order passive intermodulation of the connector under test.

[0078] A JCIMA-900-P passive intermodulation analyzer was used for testing. To ensure measurement accuracy, the instrument's intermodulation level must be below -120 dBm. Each measurement was performed in an air-conditioned room at 20 degrees Celsius to ensure temperature consistency. A torque wrench was used to minimize the effects of connector tightness. Errors during intermodulation testing were minimized by averaging multiple measurements. The instrument was warmed up for at least three minutes. The signal transmission frequency band was 930 to 960 MHz, and the intermodulation reception frequency band was 885 to 915 MHz.

[0079] The intermodulation analyzer's two signal input frequencies are 930-940MHz and 949-960MHz, while the intermodulation receiving frequency range is 885-915MHz. Therefore, the seventh-order intermodulation test results are inaccurate. By testing the magnitude of different intermodulation frequencies in swept frequency mode, we found that the magnitudes of different intermodulation frequencies for the same connector are very similar. Therefore, we changed the input signal frequencies to 933MHz and 949MHz and measured the seventh-order intermodulation magnitude at 885MHz, which we used as the seventh-order intermodulation magnitude.

[0080] Table 1 Two-tone intermodulation test results of N-type connector

[0081]

[0082] Table 1 shows the two-tone intermodulation test results of N-type connector. Figure 2 This is the spectrum diagram of the two-tone intermodulation test. Figure 2 Among them, the two highest ones in the middle are dual-tone signals, and the third-order, fifth-order, and seventh-order intermodulation products are extended to both sides.

[0083] Step 2: Test the contact resistance of the connector under test.

[0084] If the connector to be tested is an N-type connector, such as Figure 3 The figure shows a customized copper conductor that can be inserted into an N-type connector. Connecting the device at both ends of the N-type connector facilitates measuring the contact resistance of the coaxial connector. The contact resistance model is shown in FIG. Figure 4 shown.

[0085] The contact resistance model is:

[0086] The overall resistance R1 includes the contact resistance R2 of the two copper conductors and the connector, the connector contact resistance R, and the resistance R3 of the device being measured.

[0087] R1=R+2R2+R3(1)

[0088] Use a JK2511 DC resistance tester to measure the overall resistance R1. Then connect the two copper conductors and measure their contact resistance. The measurement result is considered to be the contact resistance R2 between the copper conductor and the connector. Use a vernier caliper to measure the effective length L and the cross-sectional diameter d of the N-type connector. According to the resistance determination formula (2), the N-type connection resistance R3 is calculated:

[0089]

[0090] Where ρ is the resistivity of copper;

[0091] According to formula (1), the contact resistance of the coaxial connector is:

[0092] R=R1-2R2-R3(3)

[0093] Table 2 N-type connector resistance test data

[0094]

[0095] Table 2 shows the resistance test data of N-type connector.

[0096] Step 3: Calculate the nonlinear polynomial model parameters based on the test results.

[0097] When a signal passes through a passive device, passive intermodulation (PIM) is generated due to the nonlinear characteristics of the device. To study and predict the PIM characteristics, a memoryless nonlinear polynomial model is used to describe the connector's current-voltage (IV) characteristic curve:

[0098]

[0099] Where i is the current through the connector, v is the voltage across the connector, and a k is the nonlinear polynomial model coefficient of the connector, k is an odd number;

[0100] For a two-tone CW signal, the input voltage is expressed as:

[0101] v=V[cos(2πf1t)+cos(2πf2t)](5)

[0102] Where V is the carrier amplitude, f1 and f2 are the carrier frequencies, and t represents time;

[0103] From (4) and (5), we can see that the passive intermodulation products contain odd harmonics and even harmonics. The frequency of the even harmonics is very different from the fundamental wave, while the odd harmonics are close to the fundamental wave frequency, so only the odd terms are retained. At the same time, the high-order terms of the model are very small and can be ignored. After a series of tests, the differences between retaining the model to the third order, the fifth order, and the seventh order were discussed, and retaining it to the seventh order is more appropriate. That is:

[0104] i=a1v+a3v 3 +a5v 5 +a7v 7 (6)

[0105] The nonlinear polynomial model coefficients of the coaxial connector are measured through a two-tone test experiment.

[0106] Substituting (5) into (6), we get the expression for the current, extracting the term with a frequency of 2f1-f2, and calculating the current of the third-order intermodulation product. Similarly, we can obtain the current of the fifth-order and seventh-order products:

[0107]

[0108]

[0109]

[0110] IM stands for intermodulation;

[0111] In the nonlinear model of the connector, a1 is the contact conductance, that is:

[0112]

[0113] The relationship between current, resistance and power P is:

[0114] P=i 2 R load (11)

[0115] R load is the load resistance;

[0116] Using the power of the third-order, fifth-order, and seventh-order intermodulation signals tested in step 2, the values ​​of a7, a5, and a3 are obtained according to equations (7)(8)(9)(10):

[0117]

[0118]

[0119]

[0120]

[0121] Table 3 Calculation results of nonlinear polynomial model coefficients for N-type connector

[0122]

[0123] Table 3 shows the calculation results of the nonlinear polynomial model coefficients of the N-type connector.

[0124] Step 4: Establish a mathematical model of broadband signals based on normal distribution.

[0125] To model the broadband signal, assume that the sidebands of the two broadband signals do not overlap, and each broadband signal is approximately equivalent to a CW signal with equal frequency intervals. The farther the CW signal is from the center frequency, the smaller the amplitude. According to the characteristics of broadband signals, the amplitude of broadband signals varies approximately according to a normal distribution. Assume that a broadband signal is simulated by n (n is an odd number) CW signals, and the amplitudes of the n signals are V1, V2, ..., V n , obeys the normal distribution, that is:

[0126] V~N(μ,σ 2 )(16) where μ is the average amplitude of the broadband signal, σ 2 is the variance of the amplitude, N is the normal distribution;

[0127] The power P of each CW signal in the broadband signal based on normal distribution i According to power, voltage, load resistance R load The relationship is calculated as follows:

[0128]

[0129] The signal source is set to two broadband signals with the same power and center frequencies f1 and f2. The input voltage is expressed as:

[0130]

[0131]

[0132] v=v1+v2(20)

[0133] Where v1 represents the voltage of the first broadband signal, V 1,i 、f 1,i represents the voltage amplitude and frequency of the first broadband signal, the i-th CW signal, v2 represents the voltage of the second broadband signal, V 2,j 、f 2,j Represents the voltage amplitude and frequency of the j-th CW signal of the second broadband signal.

[0134] Step 5: Establish a nonlinear polynomial mathematical model of the RF connector using the broadband signal as the excitation source, where the parameters have been calculated in Step 3; predict the power of the broadband signal passive intermodulation using the nonlinear polynomial mathematical model and analyze the results.

[0135] Substituting equation (20) in the broadband signal model into equation (6), we can obtain the expression of broadband passive intermodulation current and extract the frequency 2f 1,i -f 2,j , calculate the current of the third-order intermodulation product. Similarly, the current of the fifth-order and seventh-order products is obtained. According to the relationship between current, resistance and power, the corresponding passive intermodulation power is calculated.

[0136]

[0137]

[0138]

[0139] Using a dual-tone signal and a broadband signal based on normal distribution as the excitation source, the power of the excitation source signal is continuously changed, and the power of the third-order passive intermodulation products of the two signals is compared. The results are as follows: Figure 5 As shown in the figure, the solid line represents broadband passive intermodulation, and the dotted line represents two-tone passive intermodulation. When the power of the two signals increases from 36dBm to 42dBm, the intermodulation power difference of the N-type connector is 3.96dB at 36dBm and 3.25dB at 42dBm.

[0140] like Figure 6As shown in the figure, the power of the two-tone signal and the normally distributed wideband signal are subtracted from their respective third-order intermodulation (3OM) powers. Treating 3OM as noise, the result is called the signal-to-noise ratio (SNR). The solid line represents the SNR of the wideband signal, and the dashed line represents the SNR of the two-tone signal. As can be seen from the figure, at the same power, the SNR of the wideband signal is lower than that of the two-tone signal. As the signal power increases, the SNR of both signals decreases, and the gap between the two SNRs narrows.

[0141] Simulation of passive intermodulation circuit model based on broadband signals

[0142] Circuit modeling is a very important step in passive intermodulation spectrum analysis and power prediction. It mainly uses ADS (Advanced Design System) software to build circuit models and perform harmonic balance simulation of RF connectors.

[0143] The circuit model is mainly divided into three parts: circuit part, harmonic balance simulation, and data modulation. Figure 7 As shown in the figure, the circuit simulation diagram of the N-type connector simulated by a seventh-order polynomial model is taken as an example.

[0144] To simulate PIM under wideband signals, the wideband passive intermodulation circuit model requires the addition of more signal sources. Five single-tone signals are used to simulate a wideband signal, resulting in a total of 10 single-tone signals. This model, based on the two-tone test circuit, adds more signals, lowers the order of the signal's own harmonics to 2, and uses a wideband signal frequency interval of 2 MHz, with a normally distributed power.

[0145] The circuit model simulation model is Figure 8 The figure shows that in addition to 3rd, 5th, and 7th-order passive intermodulation, intermodulation also occurs with different frequency signals within the equivalent broadband signal. The power of each order of passive intermodulation follows a roughly normal distribution in terms of amplitude, and is somewhat lower than the two-tone test results. Data processing reveals that the sum of the power of each order of intermodulation decreases instead of increases. This indicates that using more than two carriers during harmonic balance simulation on the ADS platform affects the power of the intermodulation products. To minimize this effect, a two-tone test is performed on a single carrier of the broadband signal and a single carrier of another broadband signal. This test is repeated 25 times for all carriers. The data is then exported and the superimposed graph is plotted using Matlab.

[0146] Figure 9Using Matlab to plot the superimposed graph, this method avoids intermodulation within the equivalent broadband signal. Observing the bandwidth of each order of intermodulation, the bandwidth of the nth-order intermodulation product is n times the bandwidth of the original signal, consistent with the previous conclusion. Calculating the third-order intermodulation value in this case, the broadband signal's signal-to-noise ratio (broadband signal power minus third-order intermodulation power) is 128.0523dB. In the two-tone test, the signal-to-noise ratio is 133.0442dB, 4.3440dB higher than that of the broadband signal. This indicates that the broadband signal-to-noise ratio is worse than that of the two-tone signal, consistent with the results of the mathematical model simulation.

[0147] Use the nonlinear transfer function of the N-type connector to perform mathematical modeling and simulation. Figure 10 This is a spectrum diagram of a broadband signal model based on a normal distribution, with a power peak of 43dBm, a signal bandwidth B of 20MHz, and a frequency interval of 2MHz. The broadband signal is modified by varying B and d. Figures 11(a), 11(b), 11(c), and 11(d) show broadband passive intermodulation products at different bandwidths and frequency intervals, including third-, fifth-, and seventh-order passive intermodulation products. Because the intermodulation power near the sidebands of the passive intermodulation products is too small, only the portion above the fixed power is captured.

[0148] Figures 11(a) and 11(b) show the simulation results when the bandwidth is 20 MHz and the frequency spacing is 2 MHz and 1 MHz, respectively. The passive intermodulation of each order of the broadband signal also approximately obeys the normal distribution, and the smaller the spacing, the higher the degree of approximation. Figure 5 Comparing with Figure 11(a), we can see that the bandwidths of the 3rd, 5th, and 7th-order PIM products are 3, 5th, and 7th times the bandwidth of the original wideband signal. Considering third-order PIM as noise, the power ratio of the wideband signal to the third-order PIM is the signal-to-noise ratio. When the frequency spacing decreases from 2 MHz to 1 MHz, the signal-to-noise ratio drops from 116 dB to 110 dB, a significant decrease. This indicates that as the frequency spacing decreases, the PIM signal strength increases, leading to a decrease in the signal-to-noise ratio. Figures 11(c) and 11(d) show the simulation results for a 40 MHz bandwidth and frequency spacings of 2 MHz and 1 MHz, respectively. Comparing Figures 11(a) and 11(c) with Figures 11(b) and 11(d), we can see that increasing the bandwidth of the wideband signal increases the bandwidth of the PIM products, and the frequency overlap of each order of PIM products increases accordingly. When the bandwidth increases from 20 MHz to 40 MHz without changing the frequency spacing, the simulated data shows that the signal-to-noise ratio of the wideband signal does not change significantly.

[0149] In order to understand the difference in the behavioral characteristics of passive intermodulation products when dual-tone signals and broadband signals based on normal distribution are used as excitation sources, the nonlinear transfer function models of N, BNC, and SMA connectors are tested. Dual-tone signals and broadband signals based on normal distribution are used as excitation sources respectively. The power of the excitation source signal is continuously changed, and the power of the third-order passive intermodulation products of the two signals is compared. The results are as follows: Figure 12 As shown, the solid line represents broadband passive intermodulation, and the dashed line represents two-tone passive intermodulation. As the power of the two signals increases from 36dBm to 44dBm, at 36dBm, the intermodulation power differences for N-type, BNC, and SMA connectors are 3.96dB, 4.00dB, and 4.01dB, respectively. At 44dBm, the intermodulation power differences for N-type, BNC, and SMA connectors are 3.25dB, 3.45dB, and 3.31dB, respectively. The figure shows that broadband passive intermodulation for each connector is 3-4dB higher than two-tone passive intermodulation. The higher the signal power, the smaller the difference in the two passive intermodulation powers.

[0150] like Figure 13 As shown in the figure, the power of a two-tone signal and a wideband signal based on a normal distribution are subtracted from their respective third-order intermodulation (3OM) powers. Treating 3OM as noise, the result is called the signal-to-noise ratio (SNR). The solid line represents the SNR of the wideband signal, and the dashed line represents the SNR of the two-tone signal. As can be seen from the figure, for the same device and at the same power, the SNR of the wideband signal is lower than that of the two-tone signal. Furthermore, as the signal power increases, the SNR of both signals decreases, narrowing the gap between the two.

[0151] The above is a schematic description of the present invention and its embodiments, which is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs a structure and embodiment similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. A passive intermodulation prediction method based on normally distributed broadband signals, characterized by: The following steps are involved: Step 1: Use the two-tone intermodulation test to test the third, fifth, and seventh order passive intermodulation of the connector under test; Step 2: Test the contact resistance of the connector to be tested; Step 3: Calculate the nonlinear polynomial model parameters based on the test results; In step 3, a memoryless nonlinear polynomial model is selected to describe the current-voltage IV characteristic curve of the connector: Where i is the current through the connector, v is the voltage across the connector, and a k is the nonlinear polynomial model coefficient of the connector, k is an odd number; For a two-tone CW signal, the input voltage is expressed as: v=V[cos(2πf1t)+cos(2πf2t)] (5) Where V is the carrier amplitude, f1 and f2 are the carrier frequencies, and t represents time; The current-voltage IV characteristic curve retained to the 7th order is: i=a1v+a3v 3 +a5v 5 +a7v 7 (6) The nonlinear polynomial model coefficients of the coaxial connector are measured through a two-tone test experiment. Substituting (5) into (6), we get the expression for the current, extracting the term with a frequency of 2f1-f2, and calculating the current of the third-order intermodulation product. Similarly, we can obtain the current of the fifth-order and seventh-order products: IM stands for intermodulation; In the nonlinear model of the connector, a1 is the contact conductance, that is: The relationship between current, resistance and power P is: P=i 2 R load (11) R load is the load resistance; Using the power of the third-order, fifth-order, and seventh-order intermodulation signals tested in step 2, the values ​​of a7, a5, and a3 are obtained according to equations (7)(8)(9)(10): Step 4: Establish a mathematical model of broadband signals based on normal distribution; Step 5: Establish a nonlinear polynomial mathematical model of the RF connector using the broadband signal as the excitation source; predict the power of the broadband signal passive intermodulation through the nonlinear polynomial mathematical model and analyze the results.

2. The passive intermodulation prediction method based on a normally distributed broadband signal according to claim 1, wherein: In step 1, use a JCIMA-900-P passive intermodulation analyzer. During PIM testing, the instrument's intermodulation level must be below -120 dBm. Each measurement is performed in an air-conditioned room at 20°C to ensure temperature consistency. A torque wrench is used to minimize the effects of connector tightness. Multiple measurements are averaged to minimize intermodulation test errors. Warm up the instrument for at least 3 minutes. The signal transmit frequency range is 930 to 960 MHz, and the intermodulation receive frequency range is 885 to 915 MHz.

3. The passive intermodulation prediction method based on normally distributed broadband signals according to claim 2, wherein: In step 1, during the seventh-order passive intermodulation test, the input signal frequencies are set to 933 MHz and 949 MHz, and the magnitude of the seventh-order passive intermodulation at 885 MHz is measured.

4. The passive intermodulation prediction method based on normally distributed broadband signals according to claim 3, wherein: In step 2, if the connector to be tested is an N-type connector, the contact resistance model is: The overall resistance R1 includes the contact resistance R2 of the two copper conductors and the connector, the connector contact resistance R, and the resistance R3 of the device being measured. R1=R+2R2+R3 (1) Use a JK2511 DC resistance tester to measure the overall resistance R1. Then connect the two copper conductors and measure their contact resistance. The measurement result is considered to be the contact resistance R2 between the copper conductor and the connector. Use a vernier caliper to measure the effective length L and the cross-sectional diameter d of the N-type connector. According to the resistance determination formula (2), the N-type connection resistance R3 is calculated: Where ρ is the resistivity of copper; According to formula (1), the contact resistance of the coaxial connector is: R=R1-2R2-R3 (3).

5. The passive intermodulation prediction method based on normally distributed broadband signals according to claim 4, wherein: In step 4, the broadband signal is modeled. It is assumed that the sidebands of the two broadband signals do not overlap. Each broadband signal is approximately equivalent to a CW signal with equal frequency intervals. The farther the CW signal is from the center frequency, the smaller the amplitude. According to the characteristics of broadband signals, the amplitude of broadband signals varies according to the normal distribution. It is assumed that a broadband signal is simulated by n CW signals, and the amplitudes of the n signals are V1, V2, ..., V n , obeys the normal distribution, that is: V~N(μ,σ 2 (16) Where μ is the average amplitude of the broadband signal, σ 2 is the variance of the amplitude, N is the normal distribution; The power P of each CW signal in the broadband signal based on normal distribution i According to power, voltage, load resistance R load The relationship is calculated as follows: The signal source is set to two broadband signals with the same power and center frequencies f1 and f2. The input voltage is expressed as: v=v1+v2 (20) Where v1 represents the voltage of the first broadband signal, V 1,i 、f 1,i represents the voltage amplitude and frequency of the first broadband signal, the i-th CW signal, v2 represents the voltage of the second broadband signal, V 2,j 、f 2,j Represents the voltage amplitude and frequency of the j-th CW signal of the second broadband signal.

6. The passive intermodulation prediction method based on normally distributed broadband signals according to claim 5, wherein: In step 5, substitute equation (20) in the broadband signal model into equation (6) to obtain the expression of broadband passive intermodulation current, and extract the frequency 2f 1,i -f 2,j , calculate the current of the third-order intermodulation product. Similarly, the current of the fifth-order and seventh-order products is obtained. According to the relationship between current, resistance and power, the corresponding passive intermodulation power is calculated.

Images