A near field test device and method suitable for radio frequency device production line testing

Abstract

The application discloses a kind of near-field test equipment and method suitable for radio frequency device production line test, equipment includes equipment rack, fixture is set on rack platform, fixture is used to position device to be measured, fixture is covered in an electromagnetic shield box, electromagnetic shield box upper end opening, rack is set with X axis, Y axis walking frame above electromagnetic shield box, test arm is set with corresponding fixture on X axis, Y axis walking frame, test arm vertically extends into electromagnetic shield box from upper end opening, the front end of test arm is connected and is provided with dual-polarization antenna probe, dual-polarization antenna probe has two-way receiving antenna, two-way receiving antenna is used to simultaneously receive the radio frequency signal that device to be measured is emitted, X axis, Y axis walking frame drives test arm to be moved on device to be measured X axis, Y axis direction scanning and receiving the radio frequency signal of different positions of device to be measured, a test server is respectively connected fixture dual-polarization antenna probe and X axis, Y axis walking frame.

Description

Technical Field

[0001] This invention relates to a near-field testing device and method suitable for testing on radio frequency (RF) device production lines. The device processes the signals received from the RF test product through an algorithm to achieve stable and high-precision testing of the total radiated power (TRP) of the product's main frequency and the radiated spurious emissions (RSE) generated by the main frequency. Background Technology

[0002] Currently, with the advent of the 5G era, terminal RF chip products are gradually expanding from low-frequency narrowband coverage to high-frequency broadband coverage. The system integration and complexity are also increasing exponentially, requiring electromagnetic compatibility testing. Typically, terminal products undergo calibration testing using the 3-meter method in the laboratory stage, which requires scanning multiple dimensions with a high-gain antenna to draw conclusions. This testing process is lengthy and costly, making it unsuitable for mass production testing.

[0003] In response to this, there are currently three main batch testing solutions on the market:

[0004] 1. Reverberation Chamber-like Solution: The test is conducted by placing a circularly polarized antenna coupling plate inside a metal shielded box. The test is usually performed in the near-field radiation region. Since the signal emitted by the test is continuously reflected inside the shielded box, and the antenna coupling plate and the test are strongly coupled, even a slight offset of the coupling plate can cause a large change in the test results, introducing test errors. This causes production personnel to spend a lot of time searching for and fixing the optimal position of the coupling plate, wasting a lot of manpower.

[0005] 2. Extremely Near-Field Testing Scheme: This scheme primarily utilizes magnetic field and electric field probes to extract data from the device under test (DUT) in the near-field region. Batch data extraction is achieved through probe positioning and data acquisition systems. However, since the near-field region primarily consists of inductive field, radiation field, reactive field, and surface wave, the near-field electromagnetic field probe receives all four. To extract high-precision and stable signal data, it is typically necessary to simultaneously extract the XYZ fields using two different probes and then process the data. This process is time-consuming and involves complex algorithms. Furthermore, due to the extremely near-field extraction, there is strong mutual coupling between the DUT and the test probe, leading to significant interference. For faster testing, a method of rotating the probe 360° to find the maximum value is often used for repeated testing. The initial search time is long, and when testing batches of products, the differentiated radiation characteristics of different batches result in a low tolerance for errors and a high risk of mistesting. Additionally, due to the high-frequency efficiency limitations of the probe, it has difficulty receiving RSE signals above 10 GHz, requiring the addition of an amplifier.

[0006] 3. Multi-magnetic field probe array, similar to the principle of near-field testing, uses multiple H-plane probes to form an antenna array. The device under test is placed on the integrated board for testing. However, the position of this antenna array unit is fixed, which means that the spatial resolution cannot be modified and the probes cannot be replaced. As a result, the receiving sensitivity can only be improved by the link signal-to-noise ratio, which greatly reduces its applicability. It also has the same problems of test accuracy issues caused by positional inaccuracy and poor high-frequency receiving signals. Summary of the Invention

[0007] The purpose of this invention is to provide a high-precision near-field testing device and method suitable for production line testing. The device at the testing station employs a combination of a small dual-polarized antenna and a positioning scanning system. The received signal is processed through an algorithm to stabilize the total radiated power (TRP) of the product's main frequency and the radiated spurious emissions (RSE) generated by the main frequency.

[0008] To achieve the above objectives, the solution of the present invention is as follows:

[0009] A near-field testing device suitable for RF device production line testing includes a frame with fixtures on the frame platform for positioning the device under test (DUT). The fixtures are enclosed in an electromagnetic shielding box with an opening at the top. An X-axis and Y-axis traveling frame is mounted on the frame above the electromagnetic shielding box. Test arms are mounted on the X-axis and Y-axis traveling frames corresponding to the fixtures, extending vertically into the electromagnetic shielding box from the top opening. A dual-polarized antenna probe is connected to the front end of the test arm, which has two receiving antennas for simultaneously receiving RF signals emitted by the DUT. The X-axis and Y-axis traveling frames move the test arms along the X-axis and Y-axis directions above the DUT to scan and receive RF signals from different positions on the DUT. A test server is connected to the fixture, dual-polarized antenna probe, and X-axis and Y-axis traveling frames.

[0010] The solution further includes: the test arm is equipped with a telescopic and rotation adjustment mechanism, which connects the test arm to the dual-polarized antenna probe, thereby adjusting the distance and angle between the dual-polarized antenna probe and the device under test.

[0011] The solution is further described as follows: four clamps are set on the rack platform, the four clamps are divided into two groups, each group has two clamps, and each group of clamps is covered in an electromagnetic shielding box. The test server is connected to the dual-polarized antenna probes of the four clamps through a matrix switch.

[0012] The solution further comprises: the dual-polarized antenna probe includes two square copper-clad PCB substrates, namely substrate A and substrate B. A copper foil radiating surface with a half-open center is formed on the front of substrate A by etching, and the two sides of the half-open center have conical arc edges. A fan-shaped open-circuit microstrip line is formed on the back of substrate A by etching. Two copper foil radiating surfaces with a half-open center are formed on the front of substrate B by etching, and the two copper foil radiating surfaces have conical arc edges on the middle of the open side. A fan-shaped open-circuit microstrip line is formed on the back of substrate B. Substrate B and substrate A are connected together in a cross shape. The integrated substrate B and substrate A are fixed on the antenna base. The microstrip lines of substrate B and substrate A are respectively connected to the concentric cores of two coaxial connectors provided on the antenna base. The shells of the two coaxial connectors are connected to the metal plate of the antenna base to form a dual-polarized antenna probe with two receiving antennas.

[0013] A further aspect of the solution is that the frequency coverage of the dual-polarized antenna probe is 2-18GHz.

[0014] A near-field testing method for radio frequency (RF) device production line testing is based on the near-field testing equipment suitable for RF device production line testing, wherein: four fixtures are set on a rack platform, the four fixtures are divided into two groups, each group has two fixtures, and each group of fixtures is covered in an electromagnetic shielding box, the near-field testing method includes the following steps:

[0015] Step 1: Place four qualified standard devices in four fixtures and determine the scanning area for the devices. The scanning area is the comparison area, which includes the radiation area and the non-radiation area of ​​the signal.

[0016] Step 2: Start the device to transmit signals; start the X-axis and Y-axis traveling frame to move the dual-polarized antenna probe of the test arm to scan the scanning area on the device in the X-axis and Y-axis directions, and receive the two antenna signals of the dual-polarized antenna probe at different steps of the device scanning area in response to the transmitted signal.

[0017] Step 3: Calculate the combined amplitude of the two receiving antenna signals at different advance positions in the scanning area;

[0018] Step 4: Based on Step 3, construct a composite amplitude distribution map of different timing positions in the scanning area of ​​the standard device;

[0019] Step 5: Turn off the power and replace the standard device in the four fixtures with the device under test;

[0020] Step Six: Start the device to transmit signals; start the X-axis and Y-axis traveling frame to move the dual-polarized antenna probe of the test arm above the device and perform a scanning movement in the X-axis and Y-axis directions in the same direction as in Step One. Receive the two antenna signals from the dual-polarized antenna probe in the scanning area of ​​the device, which are positioned differently from the transmitting signal.

[0021] Step 7: Repeat step 3 to calculate the combined amplitude of the two receiving antenna signals at different advance positions in the scanning area;

[0022] Step 8: Construct a composite amplitude distribution map of different timing positions in the scanning area of ​​the device under test based on Step 7;

[0023] Step 9: Compare the distribution differences between the composite amplitude distribution map of the scanning area of ​​the standard device and the composite amplitude distribution map of the scanning area of ​​the device under test, and determine whether the device under test is qualified based on the set difference threshold.

[0024] Step 10: Turn off the power, replace the device under test, and return to Step 6 until the device under test has been tested.

[0025] The solution further includes: in step two, the minimum stepping distance in the X-axis and Y-axis directions is 0.1mm.

[0026] The solution further involves calculating the combined amplitude of the two receiving antenna signals using Formula 1.

[0027]

[0028] in:

[0029] AB is the calculated amplitude signal synthesized from the amplitude signals of the two polarized receiving antennas A and B in the two receiving antennas of the dual-polarized antenna probe;

[0030] S A With S B Based on amplitude signals A and B and phase parameter φ A φ B The constructed complex signal, where φ A φ B It is a phase quantity generated by mapping amplitude signals A and B through a nonlinear function (hyperbolic tangent tanh);

[0031]

[0032] A represents the amplitude signal received by polarized receiving antenna A in the two receiving antennas of the dual-polarized antenna probe;

[0033] B represents the amplitude signal received by the B polarized receiving antenna among the two receiving antennas of the dual-polarized antenna probe;

[0034]

[0035] K AB It is a correction factor adapted to polarized signal coupling logic;

[0036]

[0037] R AB A normalized measure representing the amplitude relationship between the two polarization channels, reflecting the impact of the amplitude difference between A and B on the relative proportion or balance.

[0038]

[0039] ∈=10^-6;

[0040] D AB It represents the relative significance of signals A and B at their own amplitude levels.

[0041]

[0042] M AB It is the average of the amplitudes of the two signals corresponding to A and B.

[0043]

[0044] The scheme further includes: the method further includes: introducing an interpolation node optimization algorithm to refine the fitting of the discretized test data for the two distribution maps respectively, and mapping and reconstructing the synthesized amplitude distribution map. The mathematical expression of the interpolation node optimization algorithm is shown in the following formula:

[0045]

[0046] in:

[0047] f^(x) is the reconstructed continuous function;

[0048] n represents the number of sampling points;

[0049] ω i This represents the weight coefficient of the i-th sampling point;

[0050] Φ() is the kernel function;

[0051] h is the bandwidth parameter;

[0052] x i These are the coordinates of the original sampling points.

[0053] The solution further includes: the method further includes quantitative analysis of the accuracy difference between the radiation zone and the non-radiation zone using an error propagation model.

[0054] Error propagation model

[0055] in:

[0056] ΔE represents the total error;

[0057] Let be the partial derivative of the interpolation function with respect to the k-dimensional parameters;

[0058] Δx k Measurement uncertainty of parameters;

[0059] When performing cubic interpolation, the error suppression effect in the radiative region is better than that in the non-radiative region. This is due to the strong correlation characteristic model of the radiation-sensitive parameters.

[0060] Strongly correlated feature model;

[0061] in:

[0062] The SSIM index is used to evaluate the structural similarity between interpolated results and true values;

[0063] μ is the mean;

[0064] σ is the covariance;

[0065] C1 and C2 are stability constants;

[0066] By introducing an interpolation node optimization algorithm, the interpolation density distribution can be dynamically adjusted using the following formula.

[0067]

[0068] in:

[0069] N opt The optimal number of nodes;

[0070] It is the second derivative of the function;

[0071] ∈ represents the preset precision threshold.

[0072] The beneficial effects of this invention are: it mainly solves the problems of low-cost, high-efficiency, and high-precision testing.

[0073] 1. Low cost: The system mainly consists of a switch matrix, dual-polarized antenna, simple aluminum profile frame, compact automatic XY axis, manual R axis, and instruments. Currently, it supports simultaneous testing of four devices simultaneously. The dual-polarized antenna can reduce coupling between products through algorithms, eliminating the need to manufacture multiple cavities and thus reducing interference between products. It also eliminates the need to find the optimal antenna position. In contrast, reverberation chambers have complex debugging processes, wasting a lot of manpower.

[0074] 2. High Efficiency: Simultaneous multi-antenna testing is possible. The scanning rack can be equipped with four antennas for simultaneous scanning, and the algorithm can increase the scanning step size by up to 5 times. Interpolation processing of the received data reduces the number of test points to one-fifth of the original number. Compared to direct testing, the processed test results show an accuracy of ±0.2dB in the radiating area and ±0.5dB in the non-radiating area. Therefore, even with slight changes in test accuracy, the number of scanned test points can be reduced to one-fifth, resulting in a one-fifth reduction in total testing time and improved efficiency. In contrast, near-field testing algorithms are more complex, struggle to remove signal interference, and are not suitable for interpolation functions.

[0075] 3. High Precision: Field extraction stability accuracy: ±0.25dB in the radiating area and ±0.5dB in the non-radiating area, supporting arbitrary rotation of the test device without changing accuracy. Because the dual-polarized antenna is a highly directional antenna with relatively stable near-field characteristics, and the energy is mainly concentrated in the direction directly in front of the antenna, only a very small portion of the energy from the sides and rear of the antenna is received. This avoids interference with the energy of the strongly radiating test device in the near field, thus not affecting the test accuracy. Furthermore, the algorithm only processes the main radiated energy received by the two main polarizations, further reducing external and surrounding interference from the test device, enabling stable multi-device testing.

[0076] The algorithm used in this invention differs from traditional methods that require both amplitude and phase information for processing in both the near and far fields of the antenna. Phase information fluctuates significantly over time and space, affecting the stability of the processed data. Furthermore, since the test area is primarily in the near field (inductive field), reflection is strong, leading to poor phase accuracy. This invention uses only amplitude information for synthesis; phase information is derived from amplitude information rather than obtained through testing. The synthesized formula displays the relative information of the total field. The inclusion of discarded phase information affects the final result, resulting in a stable and unique output value. In contrast, near-field testing equipment and magnetic field probe array solutions are highly sensitive to the state of the device under test, resulting in lower accuracy. Compared to other solutions, when multiple products are placed in the test environment at close range, severe mutual coupling occurs, making it difficult to accurately test product characteristics.

[0077] The invention will be further explained in detail below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0078] Figure 1 This is a schematic diagram of the test equipment structure of the present invention;

[0079] Figure 2 This is a schematic diagram of a dual-polarized antenna probe;

[0080] Figure 3 This is a schematic diagram comparing the synthesis of two-pole antennas;

[0081] Figure 4Amplitude distribution diagram for 10mm step synthesis;

[0082] Figure 5 Amplitude distribution diagram for 2mm step synthesis;

[0083] Figure 6 The measured amplitude distribution of a 10mm step was fitted to a 2mm composite amplitude distribution map.

[0084] Figure 7 The fitting of the measured values ​​for 10mm step and 2mm step is shown as the difference between the measured values ​​(fitting accuracy). Detailed Implementation

[0085] Example 1:

[0086] A near-field test device suitable for RF device production line testing, such as Figure 1 As shown, the near-field testing equipment includes a frame 1, which can be directly placed on the device production line. A fixture 2 is mounted on the frame platform 101 for positioning the device under test (DUT) 3. The fixture 2 is enclosed in an electromagnetic shielding box 4 with an opening 401 at the top. An X-axis and Y-axis traveling frame 5 is mounted on the frame 1 above the electromagnetic shielding box 4. The X-axis and Y-axis traveling frame 5 is driven by an X-axis drive 501 and a Y-axis drive 502. Test arms are mounted on the X-axis and Y-axis traveling frames 5 corresponding to the fixtures. 6. The test arm 6 extends vertically into the electromagnetic shielding box 4 from the upper opening 401. A dual-polarized antenna probe 7 is connected to the front end of the test arm. The dual-polarized antenna probe has two receiving antennas. The two receiving antennas are used to simultaneously receive the radio frequency signals emitted by the device under test. The X-axis and Y-axis traveling frame drives the test arm to move and scan on the device under test in the X-axis and Y-axis directions to receive the radio frequency signals at different positions of the device under test. A test server (not shown) is connected to the fixture dual-polarized antenna probe 7 and the X-axis and Y-axis traveling frame 5 respectively.

[0087] The test arm 6 is connected to a telescopic and rotational adjustment mechanism 601. The test arm 6 is connected to the dual-polarized antenna probe 7 through the telescopic and rotational adjustment mechanism 601, thereby adjusting the distance and angle between the dual-polarized antenna probe 7 and the device under test.

[0088] To improve testing efficiency: such as Figure 1 As shown, four clamps 2 are set on the rack platform. The four clamps 2 are divided into two groups, with two clamps 2 in each group. Each group of clamps is covered in an electromagnetic shielding box 4. The test server is connected to the four clamp dual-polarized antenna probes 7 through a matrix switch 8. The test server obtains the test signals of the four clamp dual-polarized antenna probes in sequence through the matrix switch 8.

[0089] The dual-polarized antenna probe in this embodiment is an antenna probe invented by the inventor, and the specific technical solution has been publicly disclosed in Chinese Patent No. 2023230658202: [e.g.] Figure 2 As shown, the dual-polarized antenna probe 7 includes two square copper-clad PCB substrates, namely substrate A 9 and substrate B 10. A copper foil radiating surface with a half-open center is formed on the front of substrate A 9 by etching, with tapered arc edges on both sides of the half-opening. A fan-shaped open-circuit microstrip line is formed on the back of substrate A by etching. Two copper foil radiating surfaces with a central opening are formed on the front of substrate B 10 by etching, with tapered arc edges on the central opening sides of the two copper foil radiating surfaces. A fan-shaped open-circuit line is formed on the back of substrate B by etching. Microstrip lines connect substrate B 10 and substrate A 9 in a cross shape to form a single unit. The integrated substrate B 10 and substrate A 9 are fixed on the antenna base 11. The microstrip lines of substrate B and substrate A are respectively connected to the concentric cores of two coaxial connectors 12 provided on the antenna base. The outer shells of the two coaxial connectors are connected to the metal plate of the antenna base to form a dual-polarized antenna probe 7 with two receiving antennas. A dust cap 13 covers substrate A 9 and substrate B 10. Other specific structures can be obtained from public documents and will not be described in detail here.

[0090] In this embodiment, the dual-polarized antenna probe covers a frequency range of 2-18 GHz. The efficiency of each frequency band of the dual-polarized antenna probe is greater than 50%, significantly higher than that of electromagnetic field probes. The probe's horizontal plane dimension is only 30 mm to minimize obstruction of the device under test (DUT). The minimum scanning step can be set to 0.1 mm, and the extraction surface has high resolution. The probe can be placed in the near-field (induction field) region, approximately 25 mm from the DUT. Since current main frequency and spurious emission testing primarily involves radiation testing, both active and passive radiation information need to be tested. Using a dual-polarized antenna, an algorithm can stably extract the radiated portion of the near-field (induction field) while simultaneously shielding interference. Furthermore, the field strength data calculated by the algorithm has a high correlation with the far-field data, allowing for the differentiation of the radiating and non-radiating areas. Compared to the complex information extracted by electromagnetic field probes, this theoretically simplifies data processing requirements. Since dual-polarized antennas are highly directional antennas with relatively stable near-field characteristics and energy mainly concentrated in the direction directly in front of the antenna, only a very small portion of the energy from the sides and rear of the antenna will be received. This will not interfere with the energy of the device under test (DUT) with strong near-field radiation and affect the test accuracy. Furthermore, the algorithm only processes the main radiated energy received by the two main polarizations, further reducing interference from external and surrounding DUTs and enabling stable testing of multiple devices simultaneously.

[0091] Example 2:

[0092] A near-field testing method for radio frequency (RF) device production line testing is based on the near-field testing equipment for RF device production line testing described in Example 1. The content of Example 1 is applicable to this example. The method includes: four fixtures are arranged on a rack platform, divided into two groups of two fixtures each, with each group of fixtures enclosed in an electromagnetic shielding box. The near-field testing method includes the following steps:

[0093] Step 1: Place four qualified standard devices in four fixtures and determine the scanning area for the devices. The scanning area is the comparison area, which includes the radiation area and the non-radiation area of ​​the signal.

[0094] Step 2: Start the device to transmit signals; start the X-axis and Y-axis traveling frame to move the dual-polarized antenna probe of the test arm to scan the scanning area on the device in the X-axis and Y-axis directions, and receive the two antenna signals of the dual-polarized antenna probe at different steps of the device scanning area in response to the transmitted signal.

[0095] Step 3: Calculate the combined amplitude of the two receiving antenna signals at different advance positions in the scanning area;

[0096] Step 4: Based on Step 3, construct a composite amplitude distribution map of different timing positions in the scanning area of ​​the standard device;

[0097] Step 5: Turn off the power and replace the standard device in the four fixtures with the device under test;

[0098] Step Six: Start the device to transmit signals; start the X-axis and Y-axis traveling frame to move the dual-polarized antenna probe of the test arm above the device and perform a scanning movement in the X-axis and Y-axis directions in the same direction as in Step One. Receive the two antenna signals from the dual-polarized antenna probe in the scanning area of ​​the device, which are positioned differently from the transmitting signal.

[0099] Step 7: Repeat step 3 to calculate the combined amplitude of the two receiving antenna signals at different advance positions in the scanning area;

[0100] Step 8: Construct a composite amplitude distribution map of different timing positions in the scanning area of ​​the device under test based on Step 7;

[0101] Step 9: Compare the distribution differences between the composite amplitude distribution map of the scanning area of ​​the standard device and the composite amplitude distribution map of the scanning area of ​​the device under test, and determine whether the device under test is qualified based on the set difference threshold.

[0102] Step 10: Turn off the power, replace the device under test, and return to Step 6 until the device under test has been tested.

[0103] The above method uses only amplitude information for synthesis, and the synthesized formula displays the total field relative information. The introduction of discarded phase information affects the final result, resulting in a stable and unique output value. Since qualified standard devices are processed using the same method, the tested products can be screened.

[0104] In step two, the minimum stepping distance in the X-axis and Y-axis directions is 0.1 mm, and the stepping distance is determined according to the test requirements.

[0105] In step three: the combined amplitude of the two received antenna signals is calculated using formula one.

[0106]

[0107] in:

[0108] AB is the calculated amplitude signal synthesized from the amplitude signals of the two polarized receiving antennas A and B in the two receiving antennas of the dual-polarized antenna probe;

[0109] S A With S B Based on amplitude signals A and B and phase parameter φ A φ B The constructed complex signal, where φ A φ B It is a phase quantity generated by mapping amplitude signals A and B through a nonlinear function (hyperbolic tangent tanh);

[0110]

[0111] A represents the amplitude signal received by polarized receiving antenna A in the two receiving antennas of the dual-polarized antenna probe;

[0112] B represents the amplitude signal received by the B polarized receiving antenna among the two receiving antennas of the dual-polarized antenna probe;

[0113]

[0114] K AB It is a correction factor adapted to polarized signal coupling logic;

[0115]

[0116] R AB A normalized measure representing the amplitude relationship between the two polarization channels, reflecting the impact of the amplitude difference between A and B on the relative proportion or balance.

[0117]

[0118] ∈=10^-6;

[0119] D AB It represents the relative significance of signals A and B at their own amplitude levels.

[0120]

[0121] M AB It is the average of the amplitudes of the two signals corresponding to A and B.

[0122]

[0123] Data processing only requires amplitude signals and not phase signals, thus eliminating phase instability in space. This formula is applicable to any angle, so there is no need to find the direction of maximum field strength of the device under test by rotating the dual-polarized antenna. The direction of the total field strength is obtained by extracting and combining the two polarization information through a switch, and the generated AB value is unique.

[0124] To improve testing accuracy, the method further includes: introducing an interpolation node optimization algorithm to refine the fitting of the discretized test data for the two distribution maps respectively, and mapping and reconstructing the synthesized amplitude distribution map. The mathematical expression of the interpolation node optimization algorithm is shown in the following formula:

[0125]

[0126] in:

[0127] f^(x) is the reconstructed continuous function;

[0128] n represents the number of sampling points;

[0129] ω i This represents the weight coefficient of the i-th sampling point;

[0130] Φ() is the kernel function;

[0131] h is the bandwidth parameter;

[0132] x i These are the coordinates of the original sampling points.

[0133] By introducing an interpolation algorithm to refine the fitting of discretized test data, the mapping and reconstruction of large-sample step test data to high-resolution parameters can be effectively achieved. Experimental results show that the interpolation accuracy of this method reaches ±0.2dB in radiation-sensitive areas and the interpolation error in non-radiation areas is controlled within ±0.5dB, significantly improving the test accuracy.

[0134] The method further includes quantifying the accuracy difference between the radiation zone and the non-radiation zone using an error propagation model.

[0135] Error propagation model

[0136] in:

[0137] ΔE represents the total error;

[0138] Let be the partial derivative of the interpolation function with respect to the k-dimensional parameters;

[0139] Δx k Measurement uncertainty of parameters;

[0140] When performing cubic interpolation, the error suppression effect in the radiative region is better than that in the non-radiative region. This is due to the strong correlation characteristic model of the radiation-sensitive parameters.

[0141] Strongly correlated feature model;

[0142] in:

[0143] The SSIM index is used to evaluate the structural similarity between interpolated results and true values;

[0144] μ is the mean;

[0145] σ is the covariance;

[0146] C1 and C2 are stability constants;

[0147] By introducing an interpolation node optimization algorithm, the interpolation density distribution can be dynamically adjusted using the following formula.

[0148]

[0149] in:

[0150] N opt The optimal number of nodes;

[0151] It is the second derivative of the function;

[0152] ∈ represents the preset precision threshold.

[0153] Figures 3 to 7 The actual measurement results are shown.

[0154] I. Algorithm Analysis Results:

[0155] The device under test is a mobile phone emitting a 2.6GHz single-tone signal. Figure 3 The two left images show the amplitude information of the two polarizations. The yellow area represents a larger radiation region, the green area a moderate radiation region, and the blue area a weak radiation region. As can be seen from the two left images, the radiation signal in the yellow area is divergent and difficult to locate. Figure 3 The right image shows the synthesis process using an algorithm to eliminate the yellow divergent signal and accurately locate the radiative and non-radiative components.

[0156] II. Fitting Analysis:

[0157] according to Figures 4 to 7 The following are the amplitude distribution diagrams of the step synthesis, where: Figure 4 Amplitude distribution diagram for 10mm step synthesis; Figure 5 Amplitude distribution diagram for 2mm step synthesis; Figure 6 The measured amplitude distribution of a 10mm step was fitted to a 2mm composite amplitude distribution map. Figure 7 The fitting of the measured values ​​for 10mm step and 2mm step is shown as the difference between the measured values ​​(fitting accuracy).

Claims

1. A near-field testing method for radio frequency device production line testing, which is a near-field testing method based on near-field testing equipment, wherein the near-field testing equipment includes an equipment rack, characterized in that, A fixture is installed on a rack platform to position the device under test (DUT). The fixture is enclosed in an electromagnetic shielding box with an opening at the top. An X-axis and Y-axis traveling frame is installed on the rack above the electromagnetic shielding box. Test arms are mounted on the X-axis and Y-axis traveling frames corresponding to the fixtures on the fixtures. The test arms extend vertically into the electromagnetic shielding box from the top opening. A dual-polarized antenna probe is connected to the front end of the test arm. The dual-polarized antenna probe has two receiving antennas, which are used to simultaneously receive the radio frequency (RF) signals emitted by the DUT. The X-axis and Y-axis traveling frames move the test arms along the X-axis and Y-axis directions above the DUT to scan and receive RF signals from different positions on the DUT. A test server is connected to the fixture, the dual-polarized antenna probe, and the X-axis and Y-axis traveling frames. Four fixtures are installed on the rack platform, divided into two groups of two fixtures each, with each group enclosed in an electromagnetic shielding box. The near-field testing method includes the following steps: Step 1: Place four qualified standard devices in four fixtures and determine the scanning area for the devices. The scanning area is the comparison area, which includes the radiation area and the non-radiation area of ​​the signal. Step 2: Start the device to transmit signals; start the X-axis and Y-axis traveling frame to move the dual-polarized antenna probe of the test arm to scan the scanning area on the device in the X-axis and Y-axis directions, and receive the two antenna signals of the dual-polarized antenna probe at different steps of the device scanning area in response to the transmitted signal. Step 3: Calculate the combined amplitude of the two receiving antenna signals at different advance positions in the scanning area; Step 4: Based on Step 3, construct a composite amplitude distribution map of different timing positions in the scanning area of ​​the standard device; Step 5: Turn off the power and replace the standard device in the four fixtures with the device under test; Step Six: Start the device to transmit signals; start the X-axis and Y-axis traveling frame to move the dual-polarized antenna probe of the test arm above the device and perform a scanning movement in the X-axis and Y-axis directions in the same direction as in Step One. Receive the two antenna signals from the dual-polarized antenna probe in the scanning area of ​​the device, which are positioned differently from the transmitting signal. Step 7: Repeat step 3 to calculate the combined amplitude of the two receiving antenna signals at different advance positions in the scanning area; Step 8: Construct a composite amplitude distribution map of different timing positions in the scanning area of ​​the device under test based on Step 7; Step 9: Compare the distribution differences between the composite amplitude distribution map of the scanning area of ​​the standard device and the composite amplitude distribution map of the scanning area of ​​the device under test, and determine whether the device under test is qualified based on the set difference threshold. Step 10: Turn off the power, replace the device under test, and return to Step 6 until the device under test has been tested. The combined amplitude of the two receiving antenna signals is calculated using Formula 1. Formula 1 in: AB is the calculated amplitude signal synthesized from the amplitude signals of the two polarized receiving antennas A and B in the two receiving antennas of the dual-polarized antenna probe; S A With S B Based on amplitude signals A and B and phase parameter φ A φ B The constructed complex signal, where φ A φ B It is a phase quantity generated by mapping amplitude signals A and B through a nonlinear function; ； ； A represents the amplitude signal received by polarized receiving antenna A in the two receiving antennas of the dual-polarized antenna probe; B represents the amplitude signal received by the B polarized receiving antenna among the two receiving antennas of the dual-polarized antenna probe; ； K AB It is a correction factor adapted to polarized signal coupling logic; ； R AB A normalized measure representing the amplitude relationship between the two polarization channels, reflecting the impact of the amplitude difference between A and B on the relative proportion or balance. ， Preset precision threshold; D AB It represents the relative significance of signals A and B at their own amplitude levels. ； M AB It is the average of the amplitudes of the two signals corresponding to A and B. 。 2. The method according to claim 1, characterized in that, In step two, the minimum stepping distance in the X-axis and Y-axis directions is 0.1 mm.

3. The method according to claim 1, characterized in that, The method further includes: introducing an interpolation node optimization algorithm to refine the fitting of the discretized test data for the two distribution maps respectively, and mapping and reconstructing the synthesized amplitude distribution map. The mathematical expression of the interpolation node optimization algorithm is shown in the following formula: ； in: The reconstructed continuous function; n represents the number of sampling points; ω i This represents the weight coefficient of the i-th sampling point; Φ() is the kernel function; h is the bandwidth parameter; x i These are the coordinates of the original sampling points.

4. The method according to claim 3, characterized in that, The method further includes quantifying the accuracy difference between the radiation zone and the non-radiation zone using an error propagation model, wherein the error propagation model is: in: ΔE represents the total error; Let be the partial derivative of the interpolation function with respect to the k-dimensional parameters; Δx k Measurement uncertainty of parameters; When performing cubic interpolation, the error suppression effect in the radiative region is better than that in the non-radiative region. This is due to the strong correlation characteristic model of the radiation-sensitive parameters, which is as follows: ； in: The SSIM index is used to evaluate the structural similarity between interpolated results and true values; μ is the mean; σ is the covariance; C1 and C2 are stability constants; By introducing an interpolation node optimization algorithm, the interpolation density distribution can be dynamically adjusted using the following formula. ； in: N opt The optimal number of nodes; It is the second derivative of the function; This is the preset accuracy threshold.

5. The method according to claim 1, characterized in that, The test arm is equipped with a telescopic and rotation adjustment mechanism. The test arm is connected to the dual-polarized antenna probe through the telescopic and rotation adjustment mechanism, thereby adjusting the distance and angle between the dual-polarized antenna probe and the device under test.

6. The method according to claim 1, characterized in that, Four fixtures are set on the rack platform. The four fixtures are divided into two groups, with two fixtures in each group. Each group of fixtures is covered in an electromagnetic shielding box. The test server is connected to the dual-polarized antenna probes of the four fixtures through a matrix switch.

7. The method according to claim 1, characterized in that, The dual-polarized antenna probe includes two square copper-clad PCB substrates, substrate A and substrate B. A copper foil radiating surface with a half-open center is etched on the front of substrate A, with tapered arc edges on both sides of the half-opening. A fan-shaped open-circuit microstrip line is etched on the back of substrate A. Two copper foil radiating surfaces with a half-open center are etched on the front of substrate B, with tapered arc edges on the middle of the open sides. A fan-shaped open-circuit microstrip line is etched on the back of substrate B. Substrate B and substrate A are connected in a cross shape. The connected substrates B and A are fixed to an antenna base. The microstrip lines of substrates B and A are connected to the concentric axes of two coaxial connectors on the antenna base. The housings of the two coaxial connectors are connected to the metal plate of the antenna base to form a dual-polarized antenna probe with two receiving antennas.

8. The method according to claim 1, characterized in that, The frequency range of the dual-polarized antenna probe is 2-18 GHz.

Images