A power divider

Abstract

The application relates to a power divider, which comprises a cavity penetrating through in the axial direction, an input port coaxial with the axis at one end, multiple output ports at the other end, a power dividing rod coaxially arranged in the cavity, an input end of the power dividing rod connected with the input port, multiple output branch ends corresponding to the multiple output ports, the power dividing rod sequentially composed of multiple impedance conversion sections, the outer diameter monotonically changed into a multi-slope segmented conical shape along the axial direction, a circumferential outer surface of each section provided with a circumferential periodical microstructure, the axial direction only covering a section length part and keeping a first predetermined clearance with a section boundary, a geometric fluctuation amount not exceeding a first predetermined proportion of a skin depth of an upper limit frequency, and a corner round transition. The power divider uniformly disperses small reflection phases, suppresses section boundary edge fields, reduces echo and standing waves, expands bandwidth, and reduces high-frequency additional loss, improves insertion loss and robustness without increasing the total length and the number of sections.

Description

Technical Field

[0001] This invention relates to passive devices, and more particularly to a power divider. Background Technology

[0002] Coaxial power dividers are widely used in RF / microwave systems such as communications, measurement and testing, and wireless infrastructure for power distribution and impedance transition between multiple ports.

[0003] Existing coaxial power dividers typically include: a hollow tubular cavity with an input side and an output side extending along its length; an input connector mounting port coaxial with the cavity axis on the input side, and several output connector mounting ports on the output side; matching connector assemblies adapted to these mounting ports. A power divider rod is coaxially arranged within the cavity, having an input end that connects to the input connector and multiple output branch ends corresponding to the multiple output connectors. The input end and each output branch end can be connected to the inner conductor via threaded holes or other means. To achieve impedance transition and power distribution, the power divider rod is generally designed with multiple impedance transformation stages along the axial direction, with each stage having different lengths and diameters to achieve progressively varying impedance matching and signal distribution.

[0004] However, existing multi-stage stepped impedance transition methods tend to cause reflected energy to accumulate at specific frequencies, resulting in large in-band echo peak-to-valley fluctuations and limited effective bandwidth. Therefore, there is an urgent need to propose a new type of power divider to solve these problems. Summary of the Invention

[0005] The purpose of this invention is to provide a power divider structure that can suppress local reflections and edge field concentrations caused by segment boundaries and achieve more uniform dispersion of reflection phase within the band without significantly increasing the total axial length and number of segments, in order to reduce echo peak value, reduce ripple and expand effective bandwidth.

[0006] The technical solution adopted by the present invention to solve the above problems is: a power divider, comprising: a cavity having a first end and a second end extending along its length; a port assembly including an input port and multiple output ports, wherein the input port is disposed at the first end and coaxially arranged with the axis of the cavity, and the multiple output ports are respectively disposed at the second end; a power dividing rod coaxially disposed inside the cavity, having an input end electrically connected to the input port and multiple output branch ends electrically connected to the multiple output ports respectively, wherein the input end is electrically connected to the input port, and each output port is connected to each of the output branch ends in a one-to-one correspondence; characterized in that: the... The power divider is composed of multiple impedance transformation segments connected sequentially along the axial direction. The outer diameter of the power divider from the input end to the output branch end varies monotonically along the axial direction and forms a segmented conical profile with multiple slopes. Microstructures are periodically distributed along the circumferential direction on the outer surface of each impedance transformation segment. The microstructures occupy a portion of the length of the impedance transformation segment in the axial direction and maintain a first preset distance between the segments and the boundaries of adjacent segments. The geometric undulation of the microstructures is limited to at most a first predetermined proportion of the electromagnetic skin depth of the conductive material of the power divider at the upper limit frequency of the corresponding operating frequency band. The corners where the microstructures meet the power divider are rounded.

[0007] The beneficial effects of the embodiments of the present invention are as follows:

[0008] By employing a segmented conical profile with a monotonically varying outer diameter along the axial direction and introducing microstructures periodically distributed along the circumferential direction on the outer surface of each impedance transformation segment, while limiting the geometric undulations of the microstructures to a first predetermined proportion of the electromagnetic skin depth at the upper limit frequency and implementing rounded transitions at corners, the phase of each level of small reflection can be more uniformly dispersed throughout the entire band, and edge field concentration and local reflection superposition at segment boundaries can be suppressed. This effectively solves the technical problems of existing multi-level stepped transitions that easily form reflection energy accumulation at specific frequency points, resulting in large echo peak and valley fluctuations and limited effective bandwidth. Thus, it achieves more continuous impedance gradient, lower standing wave and echo ripple, and wider usable bandwidth without significantly increasing the total axial length and number of segments, and reduces additional losses, improves insertion loss, and enhances structural robustness at high frequencies. Attached Figure Description

[0009] Figure 1 This is a partial cross-sectional view of a power divider shown in an embodiment of the present invention.

[0010] Figure 2 This is a partial cross-sectional view of the cavity shown in an embodiment of the present invention.

[0011] Figure 3 This is a schematic diagram of the power divider structure shown in an embodiment of the present invention. Detailed Implementation

[0012] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0013] Existing coaxial power dividers typically consist of a hollow tubular cavity 10, with an input side and an output side extending along its length. The input side has an input connector mounting port coaxial with the axis of the cavity 10, and the output side has several output connector mounting ports. Matching connector assemblies are adapted to these mounting ports. A power divider rod 30 is coaxially arranged within the cavity 10. The power divider rod 30 has an input end that connects to the input connector and multiple output branch ends corresponding to the multiple output connectors. The input end and each output branch end can be connected to the inner conductor via threaded holes or other means. To achieve impedance transition and power distribution, the power divider rod 30 is generally designed as a multi-stage impedance transformation section 310 along the axial direction, with each stage having a different length and diameter to achieve progressively varying impedance matching and signal distribution.

[0014] However, although increasing the number of stages or extending the axial length can improve the matching to some extent, existing multi-stage stepped or single-tapered impedance transitions still easily cause the small reflections between segments to converge at specific frequencies, resulting in higher in-band echo peaks, increased peak-to-valley ripples, and limited effective bandwidth. On the other hand, the edge field at the segment boundary transition is particularly concentrated at high frequencies, easily triggering the superposition of local small reflections and introducing additional losses; the lack of controlled relationship between microscale geometric features and high-frequency surface current distribution, as well as scattering management near the segment boundaries, exacerbates these problems.

[0015] Therefore, in response to the problems that existing multi-step or single-tapered transitions easily form reflection phase concentration within the band, resulting in large echo peak and valley fluctuations and limited effective bandwidth, as well as the local reflection and additional loss caused by the concentration of edge fields at the high-frequency end segment boundaries, a power divider is proposed that can achieve more uniform reflection phase dispersion and suppress the edge field at the segment boundaries without significantly increasing the total length and number of segments.

[0016] Please see Figures 1 to 3The power divider includes a cavity 10, a port assembly 20, and a power divider rod 30. The cavity 10 has a first end 110 and a first end 120 extending along its length. The port assembly 20 includes an input port 210 and multiple output ports 220. The input port 210 is located at the first end 110 and coaxially with the axis of the cavity 10. The multiple output ports 220 are each located at the first end 120. The power divider rod 30 is coaxially disposed inside the cavity 10, having an input end electrically connected to the input port 210 and multiple output branch ends electrically connected to the multiple output ports 220 respectively. The input end is electrically connected to the input port 210, and each output port 220 is connected to each of the output branch ends in a corresponding manner. The power divider rod 30 extends axially... The impedance transformation section 310 is composed of multiple sequentially connected impedance transformation sections. The outer diameter of the power divider 30 varies monotonically along the axial direction from the input end to the output branch end, forming a segmented conical profile with multiple slopes. Microstructures 320 are periodically distributed along the circumferential direction on the outer circumferential surface of each impedance transformation section 310. The microstructures 320 occupy a portion of the length of the impedance transformation section 310 in the axial direction and maintain a first preset distance between the segment boundaries of adjacent sections. The geometric undulation of the microstructures 320 is limited to at most a first predetermined proportion of the electromagnetic skin depth of the conductive material of the power divider 30 at the upper limit frequency of the corresponding operating frequency band. The corners of the microstructures 320 and the power divider 30 are rounded.

[0017] Specifically: the cavity 10 is a conductive metal shell, with a first end 110 and a first end 120 extending through it along its length. The inner wall is a continuous conductive surface, and the exterior can be fitted with a mounting and positioning part and a sealing mating part. The port assembly 20 includes an input port 210 located at the first end 110 and coaxial with the axis of the cavity 10, and multiple output ports 220 located at the first end 120. Each port uses a coaxial interface and is centered and fitted to the end face through a step surface of the shell and a positioning shoulder. The power divider 30 is coaxially arranged inside the cavity 10, having an input end electrically connected to the input port 210 and multiple output branch ends electrically connected to the multiple output ports 220. The input end is electrically connected to the input port 210, and each output port 220 is connected to its corresponding output branch end, thus forming a coaxial transmission path. The power divider 30 is composed of multiple impedance transformation sections 310 connected sequentially along the axial direction. The outer diameter of the power divider 30 varies monotonically along the axial direction from the input end to the output branch end, forming a segmented conical profile with multiple slopes. The adjacent impedance transformation sections 310 are connected by rounded corners to reduce the field strength concentration caused by geometric abrupt changes.

[0018] The reason why the outer diameter of the power divider rod 30mm is monotonically varied along the axial direction and forms a multi-slope segmented cone is firstly because the characteristic impedance of the coaxial conductor is determined by the geometry of the inner and outer conductors and the dielectric. Monotonically varying the impedance function means that it changes gradually in only one direction along the axial direction, avoiding phase aggregation and peak-to-peak superposition caused by the local reversal of "rising first and then falling". Secondly, different frequency bands have different sensitivities to the "gradual change rate". Using multiple segments with different slopes, a "gradual change" can be used in the low-frequency band to reduce the overall reflection noise floor, and a "slightly faster" slope can be used in the mid-to-high frequency band to disperse the phase of residual small reflections, achieving a more uniform in-band echo spectrum. Furthermore, without increasing the total length and number of stages, the "equivalent continuity" can be improved, the usable bandwidth can be expanded, and the echo ripple can be reduced.

[0019] The use of rounded corners between adjacent impedance transformation sections 310 is because sharp steps or small rounded corners can generate edge field peaks and current congestion at corners, leading to localized small reflections and additional conductor losses. Rounded corners significantly reduce sensitivity to minute chamfers / burrs, making the physical component closer to the design model. This reduces high-frequency additional losses, mitigates section boundary scattering, and improves insertion loss and echo stability.

[0020] Each impedance transformation segment 310 has a circumferentially distributed microstructure 320 on its outer surface. This microstructure 320 covers only a portion of the length of the impedance transformation segment 310 in the axial direction and maintains a first preset distance from the segment boundaries of adjacent segments to prevent adverse superposition of scattering at the segment boundaries and microstructure scattering. The geometric undulation of the microstructure 320 is limited to at most a first predetermined proportion of the electromagnetic skin depth of the conductive material of the power divider 30 at the upper limit frequency of the corresponding operating frequency band, thus limiting the amplitude of near-surface current disturbances. Rounded corners are used at each corner of the power divider 30 to soften the current path transitions and edge field peaks.

[0021] in:

[0022] The microstructures 320, periodically distributed along the circumference of each segment's outer circumference, possess axisymmetry. Their circumferential arrangement maintains the circumferential symmetry of the coaxial system, preventing the introduction of "azimuthally uneven" mode perturbations and avoiding energy coupling into non-target modes. Furthermore, the microstructures 320 are equivalent to introducing controlled perturbations on the conductor surface, which can passivate local electric field peaks within the segment, smooth surface current paths, and break down and separate the inevitable small reflections. Further, they can be combined with the multi-slope segmented cones corresponding to the multi-stage impedance transformation segment 310. The segmented cones perform macroscopic phase homogenization, while the microstructures 320 perform local edge field passivation; the combined effect reduces echo peaks.

[0023] The specific method for determining the first preset spacing is as follows: the segment boundary is regarded as one scattering sub-source, and the microstructure 320 is regarded as another sub-source; the "coupling sensitivity" is defined as the maximum value of the change in the equivalent reflection coefficient near the segment boundary within the band after the addition of the microstructure 320. First, an initial spacing is selected within the segment; second, the echo and insertion loss within the band are scanned, and the coupling sensitivity is recorded; then, the position of the microstructure 320 is gradually moved outward or inward closer to the segment boundary until the coupling sensitivity drops below the set threshold, while the echo peak does not rise excessively; then, it is fixed as the first preset spacing.

[0024] The purpose of the first predetermined ratio is to limit the disturbance to the skin current and the increase in conductor loss, while ensuring sufficient "passivation" capability. The specific method for determining the first predetermined ratio is as follows: First, determine threshold one, which corresponds to the loss constraint. Using the maximum allowable increase in insertion loss within the band as a constraint, sweep the fluctuations to find the upper boundary that just satisfies the insertion loss constraint; then, determine threshold two, which corresponds to the echo benefit. Without exceeding the upper boundary, select the minimum fluctuation that allows the echo peak value and peak-valley difference within the band to reach the expected value; next, threshold one and threshold two are strictly selected, and the corresponding fluctuation to skin depth ratio is the first predetermined ratio for this structural material and frequency band. Finally, verify under the disturbance model of manufacturing roughness and electroplating deviation; if the limit is exceeded, the ratio is further adjusted.

[0025] The reason for using rounded corners at each corner of the power divider 30 is similar to the reason for using rounded corners between adjacent impedance transformation sections 310.

[0026] Furthermore, the impedance transformation segment 310 is a multi-stage structure of a predetermined number, and the axial length of each stage of the impedance transformation segment 310 is limited to a second predetermined upper limit fraction based on the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band; and the ratio of the outer diameters of adjacent impedance transformation segments 310 is limited to a deviation of no more than a third predetermined upper limit ratio relative to 1 and remains monotonically changing along the axial direction to achieve continuous impedance transition.

[0027] The second predetermined upper limit fraction (the upper limit fraction whose single-stage axial length does not exceed the wavelength of the guided wave at the upper limit frequency) can be determined by the electrical length constraint method. This requires that the electrical length of each segment at the upper limit frequency be kept within the "electrically short" range, ensuring a quasi-continuous gradual change overall. Specifically: First, based on the upper limit frequency, the target echo and insertion loss indices are given. Then, the allowable single-segment phase delay upper limit is used to deduce the "electrically short" criterion. Next, this criterion is converted into a length upper limit fraction related to the guided wave wavelength, which is the second predetermined upper limit fraction. Finally, it is verified that no obvious Bragg-type standing waves or intra-segment resonance characteristics are induced across the entire band. If this is not satisfied, the fraction is further tightened. This continues until each segment is "sufficiently short," and the entire segment sequence approximates a continuous change, suppressing narrowband resonance and reflection concentration.

[0028] The deviation of the ratio of the outer diameters of adjacent segments relative to one should not exceed the third predetermined upper limit ratio and should change monotonically along the axial direction. Specifically, this means that the ratio of the outer diameters of two adjacent segments of the power divider 30 should be close to 1, i.e., a small geometric step size; and it should always "only increase" or "only decrease" along the axial direction, without any reversal. This is because a small step size makes the impedance function smoother, and the reflections at each frequency point are smaller and more dispersed; while monotony can prevent the reflection phase from being superimposed in the same direction due to local reverse inflection points. The specific determination steps are as follows: First, using the target echo peak value and peak-valley difference as constraints, gradually reduce the outer diameter step size until further reduction has little marginal benefit to the echo improvement; the deviation of the outer diameter ratio corresponding to this step size is the third predetermined upper limit ratio; then, perform a monotony check on the entire segment sequence under this ratio. If a reversal is necessary to meet assembly or strength requirements, the reversal is eliminated by reallocating the segment length and taper; then, check the in-band group delay and insertion loss to ensure that the smooth transformation does not introduce new narrowband characteristics.

[0029] The aforementioned third predetermined upper limit ratio is the tightest step upper limit that balances electrical performance and manufacturing strength to ensure continuous impedance transition.

[0030] Furthermore, the microstructure 320 consists of circumferential shallow grooves 321 periodically distributed along the circumferential direction of the power divider 30; the pitch of each circumferential shallow groove 321 is limited to a third predetermined upper limit fraction based on the wavelength of the TEM primary mode guided wave at the upper limit frequency of the operating frequency band; the geometric undulation of each circumferential shallow groove 321 is limited to at most a second predetermined proportion of the electromagnetic skin depth of the conductive material of the power divider 30 at the upper limit frequency of the operating frequency band; the groove edges of each circumferential shallow groove 321 are rounded, and the connection transition between the microstructure 320 and the power divider 30 is rounded.

[0031] Among them, the reason why the microstructure 320 is designed as an annular shallow groove is to maintain circumferential symmetry and avoid mode conversion caused by azimuth non-uniformity; and to be compatible with the field pattern of the coaxial master mode, which is equivalent to introducing axisymmetric micro-perturbation, making the influence on the advancing wave more predictable and controllable.

[0032] The method for determining the third predetermined upper limit fraction of the pitch to prevent Bragg-type reflections is as follows: using the upper limit frequency as a reference, the shallow slot period must be electrically maintained at a "subdivision" scale so that it does not form a significant periodic reflection band. The determination steps are as follows: starting from a relatively sparse pitch, gradually reduce the pitch and monitor whether periodically induced narrow band reflection peaks or microband gaps appear within the band. Once they appear, return to the position where they just disappear, which is taken as the upper limit fraction of the pitch relative to the waveguide wavelength. In addition, the influence of manufacturing limits and surface roughness on the effective pitch must also be checked at the same time, and the limit should be appropriately relaxed if necessary.

[0033] The second predetermined ratio for the groove depth undulation needs to consider both loss and return when determining it, that is, a trade-off between the set upper limit of insertion loss increment and the target of return peak value. The determination steps are as follows: with a fixed pitch, gradually increase the groove depth undulation until the return peak value is significantly reduced and the insertion loss increment approaches the allowable upper limit; take the intersection of loss constraint and return gain as the second predetermined ratio; verify under the perturbation of conductive layer thickness, coating and roughness, if the performance is sensitive to tolerance, then adjust the ratio downward. Furthermore, the circumferential shallow groove 321 provides controllable microscale passivation, the third predetermined upper limit fraction ensures "electrical subdivision", and the second predetermined ratio ensures "effective passivation within the low loss boundary".

[0034] The RF energy from input port 210 enters the cavity 10 in a coaxial fundamental mode. Through the multi-stage impedance transformation sections 310 of the power divider 30, a continuous transition from the input impedance to multiple output impedances is achieved, simultaneously distributing power among the output branches. The multi-slope segmented conical profile, through monotonically varying its outer diameter along the axial direction, further disperses the phase distribution of local reflections, thereby reducing phase concentration of reflected energy at individual frequency points over a wide frequency range. Microstructures 320 disposed around the periphery of each impedance transformation section 310 gently perturb the near-surface current path, weakening edge field peaks near the section boundaries and dispersing small reflection sources within the sections without disrupting the dominant transmission mode. The first preset distance between the microstructures 320 and the section boundaries prevents the superposition of two types of scattering. The geometric undulations of the microstructures 320 are limited by a first predetermined proportion based on the skin depth, ensuring a controlled impact on the high-frequency skin current. Combined with rounded corner transitions, this further reduces high-frequency additional losses and local echo peaks caused by geometric abrupt changes.

[0035] After the cavity 10 is machined and its inner surface treated, an input port 210 and multiple output ports 220 are installed, with their axes coaxial with the axis of the cavity 10. The power divider 30 is machined according to the geometric parameters of the segmented cone and fillet transitions. Microstructures 320 are then fabricated on the outer circumference of each impedance transformation segment 310, ensuring that the axial coverage ratio of these microstructures 320 and the first preset spacing between the segment boundaries meet design requirements.

[0036] The key elements and control points of this embodiment are: the first preset distance between the segment boundary and the microstructure 320 is used to control scattering coupling; the rounded transition of the segmented cone is used to reduce electric field abrupt changes and edge field concentration; the pitch and profile of the microstructure 320 should be consistent with the surface roughness control, and the geometric fluctuations are limited by a first predetermined ratio based on the skin depth to ensure that the influence on the near-surface current is within a controllable range.

[0037] In this embodiment, a multi-slope segmented conical profile with monotonically varying axial direction is adopted, and microstructures 320 are arranged periodically along the circumferential direction on the outer periphery of each impedance transformation segment 310, maintaining a first preset distance from the segment boundary. At the same time, the geometric undulation of the microstructures 320 is limited to a first predetermined proportion not exceeding the electromagnetic skin depth, and rounded transitions are implemented at the corners. Therefore, the local reflection phase can be uniformly dispersed within the band, and the field concentration and additional scattering at the segment boundary edge can be suppressed. This effectively solves the technical problems of high echo peak value, increased peak-valley undulation, and limited effective bandwidth caused by existing multi-level stepped or single-tapered transitions. As a result, the impedance transition is more continuous, the standing wave and echo ripple are lower, the usable bandwidth is wider, and the additional loss at the high-frequency end is reduced, while the robustness of the structure to processing and assembly deviations is improved.

[0038] Please see Figure 3 Furthermore, in some embodiments, the power divider 30 includes at least one dense region of the microstructures 320, which is a section in which the pitch density of the microstructures 320 in the axial direction of the power divider 30 is increased relative to the adjacent impedance transformation section 310; the inner wall of the cavity 10 is provided with an impedance mode suppression structure 130, which is a plurality of annular grooves 130 arrayed along the length direction on the inner wall of the cavity 10, the geometric dimensions of the annular grooves 130 are limited to a first predetermined upper limit fraction of the wavelength of the TEM primary mode guided wave at the upper limit frequency of the operating frequency band, and the impedance mode suppression structure 130 is arranged in an axial position opposite to the dense region of the microstructures 320 on the power divider 30.

[0039] Specifically: In this embodiment, the impedance mode suppression structure 130 works in conjunction with the aforementioned multi-slope segmented conical power divider 30 and circumferential periodic microstructure 320 to form a broadband low standing wave power distribution structure that suppresses reflection and mode.

[0040] The outer periphery of the power divider 30 is a support substrate with a fine microstructure 320. It can be made of a highly conductive metal, and the outer diameter sequence is set according to the target bandwidth and matching specifications.

[0041] The circumferential periodic microstructures 320 on the outer periphery of each impedance transformation segment 310 form a periodic distribution along the circumferential direction on the outer circle of each segment. The axial coverage occupies only a portion of the segment and maintains a first preset distance from the segment boundaries of adjacent segments. The microstructures 320 refine the amplitude and disperse the phase of small reflections in each segment with fine-scale geometric perturbations, while being constrained by the surface depth ratio to control additional losses. This microstructure 320 works in conjunction with the multi-slope macroscopic gradient: the multi-slope determines the full-band distribution of the reflection phase, and the microstructures 320 further refine local reflections and suppress segment boundary field spikes.

[0042] The microstructure 320 dense region refers to one or more selected segments along the axial direction of the power divider 30, where the pitch density of the microstructure 320 is increased relative to the adjacent impedance transformation segment 310, forming the microstructure 320 dense region. In axial segments with large macroscopic geometric change rates or more sensitive field energy, the ability to refine local small reflections and phase discretize is enhanced, avoiding the accumulation of single-point reflections at specific frequencies. This dense region is axially opposed to the mode suppression structure on the inner wall of the cavity 10, forming a "gate" that immediately suppresses wall modes. The dense region is integrally formed with the power divider 30, and the dense region maintains a first preset distance from the segment boundaries to avoid the superposition of segment boundary scattering. Optional materials and shapes: Same as above.

[0043] The impedance mode suppression structure 130 on the inner wall of the cavity 10 consists of an array of annular grooves 130 arranged along the length of the inner wall of the cavity 10. Each groove is circumferentially closed and axially aligned. The geometric dimensions (including the order of groove depth, groove width, and groove spacing) are limited to a first predetermined upper limit fraction based on the wavelength of the primary mode guided wave at the upper limit of the operating frequency band. This array of annular grooves 130 is positioned axially opposite the dense region of the microstructure 320. The annular grooves 130 raise the equivalent cutoff and coupling threshold of the non-primary mode through fine-scale boundary perturbations, suppressing leakage and ringing of multi-point microreflections refined from the dense region into the higher-order mode channel. The annular grooves 130 are axially opposite the dense region of the microstructure 320, constructing a mode suppression barrier at the most likely wall coupling point; a smooth inner wall is maintained outside the opposing region to preserve the low-loss channel of the primary mode. The annular grooves 130 and the cavity 10 are integrally formed by milling or electrical discharge machining, maintaining conductive continuity. Optional shapes include rectangular, trapezoidal, or rounded corner cross-sections, arranged in rows with equal or gradually varying spacing.

[0044] It should be noted that the reason for using annular grooves 130 (circumferentially closed, axially aligned) is that it allows for fine-scale modulation of the wall boundary conditions without disrupting coaxial symmetry, prioritizing the improvement of the "equivalent impedance" and "equivalent cutoff" of non-dominant modes without introducing significant narrowband Bragg reflections. Limiting the geometric scale of the annular grooves 130 to a first predetermined upper limit fraction not exceeding the wavelength of the dominant mode guide wave ensures that the microstructure exhibits "fine-scale perturbations" electrically, avoiding both over-coupling of the dominant mode and the formation of narrowband resonance. In another method for determining the first predetermined upper limit fraction, a groove-free baseline model can be established based on the upper limit frequency of the operating band. The groove scale is gradually increased, and the in-band echo, insertion loss, and the appearance of narrowband reflection peaks are observed. The normalized scale is "frozen" near the optimal combined value of in-band echo and insertion loss without triggering narrowband peaks, serving as the first predetermined upper limit fraction. Subsequently, manufacturing and assembly disturbances are superimposed for robustness verification. If sensitive to disturbances, the scale is appropriately lowered, taking the strictest of the multiple constraints.

[0045] The reason why the annular groove 130 is axially opposed to the dense area of ​​the microstructure 320 is that the dense area of ​​the microstructure 320 decomposes the local reflection into multi-point micro-reflection and generates more near-field components of the wall. The mode suppression groove is set in this area opposite to the inner wall, which can shorten the propagation path of micro-reflection coupling to the wall, raise the non-primary mode threshold in place, and prevent these micro-perturbation energies from being carried by higher-order modes and forming ringing in the cavity. Together with the macroscopic multi-slope gradient and segment boundary rounded corner of the power divider 30, it achieves the dual constraint of "refining reflection and suppressing modes".

[0046] During operation, the power divider receives the input signal in coaxial master mode via input port 210 into cavity 10. Macroscopic impedance gradients are achieved on the multi-slope segmented cone of the power divider rod 30. The circumferential periodic microstructures 320 on the outer periphery of each segment refine and phase-dispersegate unavoidable small reflections within their coverage area. Furthermore, due to maintaining a first preset distance from the segment boundaries and rounded corners, field spikes at the segment boundaries are weakened. Upon reaching the dense region of the microstructures 320, this region further enhances the refinement and phase dispersion of microreflections. The opposing inner wall annular mode-suppression groove array improves the equivalent cutoff of non-master modes, forcing wall-coupled components to be difficult to propagate or reside in higher-order modes. Microreflections attenuate or cancel each other out within the master mode frame. The signal ultimately achieves power distribution across multiple ports according to the design, with in-band echo and ripple convergence, and the master mode channel maintaining low loss.

[0047] Key aspects of the power divider in this embodiment include: Transient start-stop: Opposite relationships limit transient scattering accumulation on the wall surface; multiple slopes and rounded corners reduce startup spike echoes. Steady-state broadband: Macroscopic multiple slopes determine the full-band phase distribution; the microstructure 320 and dense areas refine the reflection amplitude; mode-suppression grooves raise the non-dominant mode threshold; the combination of these three factors causes echo peaks and valleys to converge. High-power temperature rise: The grooves must ensure continuous conductivity and a good heat dissipation path; the surface quality of the microstructure 320 is stable to control additional losses.

[0048] In particular, during manufacturing and assembly, coaxial positioning and axial reference fixtures ensure that the dense area—the groove array—is aligned; the fillets and the first preset spacing reduce the sensitivity to small tolerances.

[0049] In this embodiment, by employing a multi-slope segmented conical power divider 30 in conjunction with circumferential periodic microstructures 320 on the outer periphery of each segment, and by arranging annular mode-suppressing grooves in the dense area of ​​the microstructures 320 of the power divider 30 and on the inner wall of the cavity 10, and by designing the geometric dimensions of the grooves with constraints, the problem of reflection aggregation at specific frequency points and leakage to higher-order modes on the cavity wall caused by multi-level stepped transitions is effectively solved, resulting in echo peak and valley fluctuations and additional losses. Thus, the technical effect of obtaining more continuous impedance gradient smooth in-band echo and more stable low-loss master mode transmission is achieved without increasing the total axial length and the number of segments.

[0050] Please see Figure 3In one embodiment, the geometry of the annular groove 130 includes groove depth, groove width, and groove spacing; and is respectively limited to a fourth predetermined upper limit fraction, a fifth predetermined upper limit fraction, and a sixth predetermined upper limit fraction based on the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band; the width of the impedance trench 140 is limited to a corresponding seventh predetermined upper limit fraction relative to the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band.

[0051] Specifically: This embodiment relates to a geometric dimension limitation of the mode suppression structure and port impedance boundary structure in a coaxial power divider, which is suitable for controlling the propagation capability of higher-order modes in the cavity 10 and stabilizing the isolation characteristics between the output ports 220.

[0052] The annular groove 130 is a circumferential recess feature set on the inner wall of the cavity 10. Multiple annular grooves 130 are distributed sequentially along the axial direction of the cavity 10, forming an array along the length direction. Each annular groove 130 has three geometric parameters: groove depth, groove width, and groove spacing. The groove depth is the depth to which it is recessed radially into the inner wall of the cavity 10; the groove width is the dimension occupied by the groove along the axial direction; and the groove spacing is the axial interval between two adjacent annular grooves 130. The array of annular grooves 130 forms a periodic boundary perturbation on the inner wall of the cavity 10, making the inner wall of the cavity 10 present a discontinuous conductive boundary for higher-order modes above the main transmission mode, thereby improving the equivalent cutoff characteristics of these higher-order modes. The groove depth determines the perturbation intensity of the groove on the current path of the cavity wall, the groove width affects the distribution of this perturbation in the axial direction, and the groove spacing determines the degree of linear induction or coupling between the grooves in the axial direction. When the groove depth, groove width, and groove spacing are all kept within a predetermined small fraction of the wavelength of the main transmission mode guided wave at the upper limit frequency of the operating frequency band, the influence of the annular groove 130 on the main transmission mode remains at a perturbation level, while it plays a significant cutoff role for higher-order modes that are not desired to propagate. The annular groove 130 is arranged along the inner wall of the cavity 10, facing the radiation side of the power divider 30, which is equivalent to introducing an axially distributed "modal filter band" in the medium space between the power divider 30 and the cavity 10. This space is the channel through which electromagnetic field energy is mainly distributed, so the groove is directly coupled to this channel. The annular groove 130 can be obtained by direct cutting or forming of the inner wall of the cavity 10 without additional assembly. The annular groove 130 can be an approximately rectangular cross-section, an approximately trapezoidal cross-section, or a concave cross-section with rounded corners, as long as the groove depth, groove width, and groove spacing can still be defined.

[0053] The slot depth, slot width, and slot spacing are limited to no more than the fourth, fifth, and sixth predetermined upper limit fractions, based on the wavelength of the main transmission mode guided wave at the upper limit frequency of the operating frequency band, as detailed below:

[0054] The above-mentioned requirements stipulate that the three geometric dimensions of slot depth, slot width, and slot spacing should all be kept within their respective small fractional ranges relative to the waveguide wavelength of the main transmission mode at the upper limit frequency of the operating frequency band, rather than being arbitrarily chosen.

[0055] The guided wave wavelength corresponding to the upper limit of the operating frequency band is the shortest target transmission wavelength in the system. At this scale, if the groove depth, width, and spacing are all kept to be a small fraction of the guided wave wavelength, the influence of each groove on the main transmission mode can be considered a minor perturbation. It will not form a strongly coupled standing wave structure on the path of the main transmission mode, nor will it be equivalent to an independent resonant cavity. On the other hand, for higher-order modes, this set of periodic perturbations will disrupt their continuous propagation conditions, making it difficult for higher-order modes to form stable transmission in this cavity region. In other words, the smaller the groove depth, width, and spacing, the more effectively it provides a "dense and shallow" periodic mode-suppression filter band, locking out higher-order modes without destroying the main mode.

[0056] The fourth predetermined upper limit fraction corresponds to the slot depth. The slot depth determines the disturbance intensity of the wall current. If the slot depth accounts for too large a proportion of the guided wave wavelength, the slot may form a visible equivalent cavity, which in turn introduces new parasitic resonances. The fourth predetermined upper limit fraction is set by simulating the proportion of the slot depth and observing the changes in the echo of the main transmission mode and the cutoff margin of higher-order modes. The minimum upper limit that prevents a sudden increase in the echo of the main mode and significantly weakens the propagation of higher-order modes is taken as this fraction.

[0057] The fifth predetermined upper limit fraction corresponds to the slot width. The slot width affects the spatial expansion of a single slot in the axial direction. If the slot width is close to a large proportion of the guided wave wavelength, a single slot will begin to behave as a standing wave cavity segment along the axial direction, which may couple with the dominant mode. The fifth predetermined upper limit fraction is determined by scanning different slot width proportions through simulation, selecting the maximum allowable width proportion that can maintain the high-order mode suppression capability without significantly deteriorating the dominant mode transmission loss and return.

[0058] The sixth predetermined upper limit fraction corresponds to the slot pitch. The slot pitch determines the periodic distribution pitch between adjacent slots. If the slot pitch is close to a large proportion of the wavelength of the dominant mode guide wave, these slots may form Bragg-type reflection bands in the higher-order mode frequency band, producing a strong narrowband effect or even modulating the dominant mode. Conversely, when the slot pitch is limited to a small fraction, the array exhibits a gentle, continuous rough boundary for the dominant mode, but still disrupts the continuous path for higher-order modes. The sixth predetermined upper limit fraction can be determined by performing field distribution analysis on different slot pitches, selecting the maximum allowable pitch proportion that will not produce strong narrowband reflection peaks at the upper limit frequency of the operating frequency band but still reduces the field strength of higher-order modes.

[0059] Through the above restrictions, the inner wall of cavity 10 will not form an additional equivalent resonant cavity in the main transmission mode path, while still preventing the energy of higher-order modes from effectively accumulating along the length of cavity 10. The slot depth, slot width, and slot spacing can vary slightly in different axial sections, but each location must still maintain a limit not exceeding the corresponding fourth, fifth, and sixth predetermined upper limit fractions.

[0060] The width of the impedance trench 140 is limited to no more than the corresponding seventh predetermined upper limit fraction, as follows:

[0061] Impedance trench 140 is an annular trench disposed at the end face of output port 220 along the circumferential direction. The trench has an opening width. The width is limited to a seventh predetermined upper limit fraction of the wavelength of the main transmission mode guided wave at the upper limit frequency of the operating frequency band.

[0062] The purpose of the impedance trench 140 is to establish a near-virtual ground boundary condition near the port opening plane, suppressing coupling between ports. Its electrical behavior approximates a resonant impedance joint of a quarter-waveguide wavelength. When the trench opening width is too large, the trench may form an excessively strong radiation or coupling port for the main transmission mode, leading to increased insertion loss or excessively large local divergent electric field at the end face, disrupting the matching. By keeping the width a small fraction of the main transmission mode's waveguide wavelength, the trench primarily exhibits approximate deep resonance characteristics rather than a large-opening radiation slot, thus introducing a high-suppression boundary around the port plane without significantly sacrificing the main mode transmission.

[0063] The seventh predetermined upper limit score is determined by simulating different trench width ratios at the ports, recording port isolation, port return ratio, insertion loss, and local field strength distribution at the ports. When the trench width increases to the point where end-face radiation leakage becomes significant, port return deteriorates, or port isolation no longer improves, the critical upper limit of the waveguide wavelength ratio corresponding to the trench width is taken as the seventh predetermined upper limit score.

[0064] Impedance trench 140 mainly functions near the port plane, and has its own function as an annular groove 130 on the inner wall of cavity 10. The former establishes an approximate virtual ground boundary on the end face, while the latter enhances the cutoff of higher-order modes inside cavity 10. The two work together to suppress crosstalk channels between ports and parasitic modes inside cavity 10.

[0065] When the power divider is operating, the main transmitted energy is transmitted to the output port 220 along the power divider rod 30 in the form of the main transmission mode. An array of annular grooves 130 is located on the inner wall of the cavity 10. Through controlled small-scale perturbations of groove depth, width, and spacing, higher-order modes are made difficult to form stable propagation within the cavity 10, thereby reducing the residence time and coupling path of parasitic modes within the cavity. Simultaneously, an impedance trench 140 is located on the end face of the output port 220. The trench width is limited to a predetermined upper limit fraction, making the trench more like a longitudinal resonant impedance joint rather than a radial opening. This establishes an approximate virtual ground boundary in the port plane region, blocking lateral leakage channels between ports through the flange face or shell surface. This achieves the combined effect of the inner cavity propagation section and the end face opening section: the inner cavity section is unfavorable for the continued outward transmission of higher-order modes, and the port section is unfavorable for the lateral coupling of residual energy at the end.

[0066] Under steady-state broadband operating conditions, the annular groove 130, through its limited groove depth, width, and spacing, suppresses the high-order mode energy carrying path at the inner wall of the cavity 10, thus suppressing ringing within the cavity. Under high power or high frequency conditions, this suppression limits the accumulation of high-field hotspots on the inner wall of the cavity 10, reducing additional losses caused by local high fields. The impedance trench 140 at the output port 220 maintains an approximate virtual ground boundary through its limited width. Even if there is external load mismatch or high reflection at a single port, the impedance trench 140 still suppresses the mismatch energy from transgressing laterally to other ports, improving the decoupling stability between ports. If the operating conditions are close to the upper limit of the operating frequency band, the guided wave wavelength is the shortest, and the fractional constraints are the most stringent, preventing new narrowband parasitic resonances from appearing at the highest frequency end.

[0067] In this embodiment, by employing annular grooves 130 arranged axially on the inner wall of the cavity 10 and limiting the groove depth, width, and spacing to no more than a fourth, fifth, and sixth predetermined upper limit fraction based on the wavelength of the main transmission mode guided at the upper limit frequency of the operating frequency band, and by forming impedance trenches 140 on the end face of each output port 220 and limiting the width of the impedance trenches 140 to no more than the corresponding seventh predetermined upper limit fraction, the technical problem of insufficient port isolation, increased in-band crosstalk, and high VSWR peak at high frequencies is effectively solved in the traditional structure, where higher-order modes are easily allowed to propagate on the wall surface inside the cavity 10 and form coupling leakage paths between the output ports 220. This results in insufficient port isolation, increased in-band crosstalk, and high VSWR peak at high frequencies. Thus, the technical effect of simultaneously weakening the propagation of parasitic modes inside the cavity and constructing a decoupling boundary at the port plane is achieved in a wide frequency band, thereby obtaining stable isolation, low crosstalk, and low VSWR.

[0068] Furthermore, in some embodiments, the axial length of each impedance transformation segment 310 is configured such that the difference between the integrals of the electromagnetic energy density along the axial direction of the power divider 30 of adjacent impedance transformation segments 310 in the dielectric space between the power divider 30 and the cavity 10, based on the field distribution at the upper limit frequency of the operating frequency band, does not exceed a second predetermined upper limit ratio based on the larger of the two integrals.

[0069] Specifically, in this embodiment, the power divider rod 30 is configured with an impedance transformation section 310 of equal field energy segmentation, which is used to determine the axial length of each segment in the coaxial power divider according to the principle of energy balance, so as to obtain broadband low standing wave and low loss power distribution.

[0070] The power divider 30 is composed of multiple sequentially connected impedance transformation segments 310 along its axial direction, with the downstream segment of any segment designated as its adjacent segment. The impedance transformation segments 310 achieve an overall impedance transition through a gradual geometric change. Each impedance transformation segment 310 is located within a coaxial dielectric space between the outer circumference of the power divider 30 and the inner wall of the cavity 10, and they are smoothly connected to each other to ensure continuous electromagnetic field transmission. The multi-stage impedance transformation segments 310 are integrally machined or formed as an equivalent sequence to create a continuous cross-section.

[0071] The medium space is an annular volume enclosed by the axial projections of the outer surface of the power divider 30 and the inner wall of the cavity 10. Its function is to provide energy storage and propagation paths for the master mode electromagnetic field. Its function is to carry the transmission of input power to each output branch.

[0072] The field distribution at the upper limit of the operating frequency band (reference field distribution) refers to the steady-state three-dimensional electromagnetic field intensity distribution obtained by excitation of the port with the main transmission mode at the highest frequency point of the specified operating frequency band. Its function is to provide a unified scale for energy density assessment within a segment, and to use this distribution as the evaluation benchmark when solving for the length of each segment.

[0073] The reason for using the upper limit frequency as the benchmark is that the upper limit frequency corresponds to a shorter wavelength and stronger geometric sensitivity, and reflection and field spikes are more likely to occur at the segment boundary; checking the energy balance under this condition can cover more stringent conditions in the whole band and ensure that the low frequency end is naturally satisfied.

[0074] The axial integral difference constraint of electromagnetic energy density refers to integrating the electromagnetic energy density along the axial direction within the dielectric space of any pair of adjacent segments to obtain the energy integral values ​​of the two segments. The larger of the two values ​​is used as the normalization benchmark, and the difference must not exceed a second predetermined upper limit ratio. The purpose of the axial integral difference constraint of electromagnetic energy density is to ensure that adjacent segments bear approximately equal amounts of field energy and reflection contribution, so that small reflections are evenly distributed along the axial direction, avoiding echo peaks caused by overload in a certain segment. The axial integral difference constraint of electromagnetic energy density serves as a decision criterion for solving and correcting segment lengths, constraining the axial length of each segment.

[0075] The second predetermined upper limit ratio is a normalized difference threshold set to ensure energy balance, and its value is less than one. Its function is to provide an acceptable upper limit for energy imbalance between adjacent segments, used for segment length iteration stopping conditions and tolerance robustness verification. The method for determining the second predetermined upper limit ratio is as follows: First, establish a three-dimensional model of the device, set conductive boundaries and port matching, and select the highest frequency point of the operating frequency band for field solution; second, using the current segmentation scheme as the initial value, calculate the axial integral of the energy density in the dielectric space of each segment; next, compare the normalized difference between adjacent segments. If it exceeds the second predetermined upper limit ratio, shorten the segment with higher energy or lengthen the segment with lower energy, keeping the overall monotonic profile unchanged; then, repeat the field solution and integration until all adjacent segments meet the threshold; finally, perform a sensitivity check under disturbances such as manufacturing and assembly tolerances and temperature drift. If the limit is exceeded, tighten the threshold and fine-tune the segment length.

[0076] The technical solution of this embodiment provides macroscopic gradation and local micro-perturbation refinement with the aforementioned multi-slope segmented cone and circumferential micro-structure 320. This embodiment determines the segment length based on the principle of energy balance, so that the refined micro-reflections are evenly distributed in the axial direction. The aforementioned inner wall annular mode suppression groove is arranged in the opposite position of the sensitive area to further block the bearing and ringing of higher-order modes on the wall surface. The three constitute a closed-loop synergy of geometric gradation, energy balance and modal gating.

[0077] During power divider operation, input power is injected through input port 210, propagates axially within the dielectric space enclosed by power divider rod 30 and cavity 10, and is distributed to each output branch. A multi-slope profile provides a continuous impedance transition, and the circumferential microstructure 320 refines and disperses the inevitable small reflections within the segment. The segment length, determined according to the energy balance principle, ensures near-equilibrium energy distribution between adjacent segments, preventing small reflections from concentrating at individual segment boundaries axially. Opposite annular grooves 130 on the inner wall raise the equivalent cutoff threshold of the non-dominant mode, limiting modal ringing caused by wall coupling. Ultimately, the dominant transmission mode advances with low loss axially, in-band echo peaks and valleys converge, and insertion loss remains at a low level.

[0078] In this embodiment, the power divider can balance energy during start-up and shutdown transients, preventing sudden field spikes in individual segments at the moment of power-on. Furthermore, the equal-field energy segmentation at high frequencies makes the reflection contribution across the entire band more uniform. When port mismatch occurs or power increases, the balanced segmentation reduces the risk of overload in a single segment, and, in conjunction with the inner wall mode-suppressing structure, reduces the duration of ringing within the cavity.

[0079] In this embodiment, by employing an axial length configuration technique based on equal field energy segmentation with controlled energy integral difference between adjacent segments, the technical problem of reflection contribution gathering at individual segment boundaries in multi-segment geometric gradients, leading to in-band echo peak and valley fluctuations and increased additional loss, is effectively solved. This achieves the technical effect of balanced reflection distribution and stable low-loss transmission in the main channel.

[0080] Please see Figure 1 Furthermore, in some embodiments, the end face of each output port 220 forms an impedance trench 140 along the circumferential direction; the geometric depth of the impedance trench 140 is configured such that the equivalent electrical length of the impedance trench 140 at the upper limit frequency of the operating frequency band approaches one-quarter of the guided wave wavelength of the TEM master mode; and the impedance trench 140 and the impedance mode suppression structure 130 maintain a second preset distance in the axial direction.

[0081] Specifically: This embodiment discloses a collaborative mode suppression and isolation structure of port end face impedance trench 140 and cavity inner wall impedance mode suppression structure 130, which is suitable for improving port isolation, reducing high-order mode coupling and stabilizing port matching at the output port 220 of a coaxial power divider.

[0082] Output ports 220 are the various power output interfaces installed at the first end 120 of the power divider. Their end faces are conductive open planes surrounding the output interface facing the external system, typically the mating area between the conductor termination surface and the external RF connector. This end face is both the location where the main transmitted energy is coupled to the external load and a sensitive location where internal field energy may radiate outwards or cause crosstalk between ports. Output ports 220 are electrically connected to the corresponding output branch of the power divider 30, ensuring stable coupling of the main transmitted mode energy from the power divider 30 to the outside. Output ports 220 are typically fixed at the mounting opening at the first end 120 of the cavity 10, ensuring conductive continuity between the outer conductor of the port and the cavity 10, and conductive connection between the inner conductor and the output branch of the power divider 30. The end face can be integrally machined from a highly conductive metal or formed from a connector flange surface, maintaining conductive continuity with the cavity 10 as a whole.

[0083] Impedance trench 140 forms one or more closed annular recesses along the circumferential direction on the end face of each output port 220. This trench surrounds the port opening area, presenting an annular cavity structure recessed into the cavity 10, belonging to the geometric groove within the conductive wall. Electromagnetically, the impedance trench 140 manifests as a resonant impedance structure of a specific electrical length, providing an approximate virtual ground boundary for specific modes at the port (i.e., high reflection and low penetration for common-mode or coupled energy), thereby weakening the energy coupled between ports through the surface of the cavity 10. In other words, it artificially establishes a highly suppressed boundary condition at the port plane. The impedance trench 140 surrounds the output port 220, acting directly near the port opening, rather than being located inside the main propagation channel of the power divider 30; unlike the mode-suppressing structure on the inner wall of the cavity 10, the impedance trench 140 primarily suppresses coupling leakage paths near the port plane. The impedance trench 140 can be directly machined into the port metal surface by turning, milling, or other means, without relying on additional components for fixation. The trench wall and the port metal surface maintain electrical continuity. A single annular groove or multiple concentric annular grooves can be used, as long as they maintain the form of surrounding the output port 220 in a circular direction.

[0084] The reason why the geometric depth of the impedance trench 140 is configured so that its equivalent electrical length approaches one-quarter of the guided wave wavelength is because it is designed based on the effective electrical length of the equivalent transmission line segment or resonant cavity segment formed by the impedance trench 140 in the radio frequency sense, so that the equivalent electrical length is close to one-quarter of the guided wave wavelength of the main transmission mode at the upper limit of the operating frequency band. When the effective electrical length of a transmission structure is close to one-quarter of the guided wave wavelength of the main mode, the structure, with one end open and the other end approximately closed, will exhibit a boundary characteristic close to high or low impedance at the open position. Adjusting this characteristic to the output port 220 plane can artificially create an approximate "virtual ground" or high isolation boundary near the port plane, making it difficult for coupled energy to leak between ports.

[0085] TEM Master Mode: In coaxial power dividers, the main transmission mode is an approximate transverse electromagnetic mode, meaning the electric and magnetic fields are primarily distributed laterally, with no longitudinal field component dominating the transmission in the axial direction. This mode corresponds to the basic transmission mode of a conventional coaxial line and is often referred to as the TEM master mode. This TEM master mode has a defined waveguide wavelength, meaning the main energy propagates in this structure with a phase constant determined by this wavelength. It should be noted that the depth of the impedance trench 140 is not simply a matter of "the deeper the better" or "the shallower the better," but rather adjusted so that its equivalent electrical length reaches approximately one-quarter of the waveguide wavelength of the TEM master mode at the upper frequency limit. Once this condition is met, the trench strongly suppresses the coupling path of the master mode, significantly reducing energy crosstalk between the output ports 220. The equivalent electrical length can be fine-tuned by slightly altering the trench opening width, bottom transition fillet, or trench sidewall slope, without relying entirely on a single physical depth. The TEM master mode is the power transmission path that the system intends to retain. All of the aforementioned structures (including the impedance trench 140 and the mode suppression groove) are designed to control energy leakage and parasitic mode coupling outside the TEM master mode, ensuring that the final output is still dominated by the TEM master mode, rather than introducing parasitic resonances of higher-order modes.

[0086] The second preset distance between the impedance trench 140 and the impedance mode suppression structure 130 is as follows: the impedance trench 140 is located near the end face of the output port 220; the impedance mode suppression structure 130 (i.e., the annular grooves 130 arrayed along the length direction on the inner wall of the cavity 10) is located in the main cavity propagation region. The two maintain a non-zero distance in the axial direction, which is the second preset distance. This second preset distance allows the local virtual ground effect of the port opening plane (generated by the impedance trench 140) and the mode suppression structure in the main cavity (generated by the annular grooves 130) to improve the cutoff of higher-order modes to function in different effective ranges, preventing the two structures from directly coupling into a new parasitic resonance path. In other words, the impedance trench 140 is responsible for "sealing" the port face, and the mode suppression structure is responsible for "improving the cutoff" in the propagation region of the cavity 10. The two work together to suppress inter-port coupling and re-excitation of higher-order modes within the cavity through spatial isolation. The existence of the second preset spacing ensures that energy passes through two "gates" before leaving the main propagation region and reaching the port plane: firstly, the higher-order modes of the main cavity are difficult to maintain propagation; secondly, the port opening is approximately a virtual ground, preventing residual energy from leaking to adjacent ports. The second preset spacing can be achieved by selecting the axial position of the anti-mold groove on the inner wall of the cavity 10, or by making appropriate adjustments to the thickness of the port flange.

[0087] In operation, radio frequency energy is coupled to the corresponding output port 220 via the output branch of the power divider 30. The main transmission mode propagates outward along the coaxial path. At this time, the impedance trench 140 generates a resonant boundary with an equivalent electrical length of approximately one-quarter of the guided wavelength around the opening surface of the output port 220. This effect is similar to forming a high suppression region or an approximately virtual region near the port plane, thereby limiting the lateral diffusion of the electromagnetic field along the metal surface of the cavity 10 or the port flange region to other output ports 220. Meanwhile, in the inner main cavity region, the impedance mode suppression structure 130 improves the equivalent cutoff of higher-order modes through an array of annular grooves 130 along the inner wall, making it difficult for non-target modes to maintain propagation within the cavity. When energy attempts to form a parasitic loop between the cavity wall and the port, these two structures exert a dual suppression effect by dividing their functions: the main cavity section is unfavorable for higher-order modes to continue forward, and the port section is unfavorable for residual coupling to diffuse to adjacent ports at the opening.

[0088] Under normal steady-state operation, each output port 220 feeds power to an external load. The impedance trench 140 significantly weakens the crosstalk path between ports, improving the isolation between adjacent ports. At high-frequency or high-power ends, the impedance mode suppression structure 130 simultaneously limits the residence time of higher-order modes, reducing the possibility of repeated ringing within the cavity. In power-on transients or port mismatch conditions, the impedance trench 140 forms an approximate virtual point near the port plane, causing mismatch energy to be more likely to be locally reflected near the port rather than coupled to other ports, reducing mutual interference. The second preset spacing ensures that these two types of operating regions are not strongly coupled, avoiding the formation of new narrowband parasitic resonances.

[0089] In this embodiment, by employing the technical means of forming impedance trenches 140 along the circumferential direction on the end face of each output port 220 and configuring the geometric depth of the impedance trenches 140 such that the equivalent electrical length at the upper limit frequency of the operating frequency band approaches one-quarter of the guided wave wavelength of the main transmission mode, and maintaining a second preset distance between the impedance trenches 140 and the impedance mode suppression structure 130 located on the inner wall of the cavity 10 in the axial direction, the technical problems of insufficient port isolation caused by coupling between the output ports 220 through the wall surface of the cavity 10 and the port plane, as well as crosstalk and high standing wave peak caused by the easy re-excitation of higher-order modes in the cavity, are effectively solved. Thus, the technical effect of simultaneously suppressing crosstalk between ports and parasitic modes in the cavity and improving isolation and stable output matching characteristics under wide operating frequency band conditions is achieved.

[0090] Please see Figure 1 and Figure 3In some embodiments, the power divider further includes a slotted cylindrical temperature-sensitive dielectric sheath 40 and a circumferential resistive film 50. The temperature-sensitive dielectric sheath 40 is disposed on the outer surface of the power divider rod 30 and located outside a compensation section determined by a temperature sensitivity criterion along the axial direction of the power divider rod 30. The temperature sensitivity criterion is based on the impedance and / or phase shift with temperature at the upper limit frequency of the operating frequency band reaching a first predetermined threshold. The temperature-sensitive dielectric sheath 40 is made of a dielectric material with a non-zero dielectric constant temperature coefficient. The circumferential resistive film 50 is disposed outside the temperature-sensitive dielectric sheath 40 and is only disposed within a low-field annular band determined by a node criterion along the axial direction of the power divider rod 30. The node criterion is based on the local electric field amplitude normalized to the preceding wave component at the upper limit frequency of the operating frequency band not exceeding a fourth predetermined upper limit ratio. The circumferential resistive film 50 has non-conductive segmented slits in the circumferential direction and a gradually tapered end at its axial edge.

[0091] Specifically: This embodiment relates to an additional structure for a power divider that includes a temperature-sensitive dielectric sheath 40 and a circumferential resistive film 50. This additional structure is used to simultaneously achieve temperature drift compensation and fixed-point reflection energy dissipation on the outside of the power divider rod 30, so as to improve broadband stability, isolation and environmental robustness.

[0092] The temperature-sensitive dielectric sheath 40 is a slotted cylindrical sleeve, hollow in shape, with a circumferential covering feature, and is fitted along the outer circumference of the power divider 30. The sleeve has circumferential slots to prevent the formation of a fully closed conductor or dielectric ring, which could lead to parasitic circulating currents or resonances, and also facilitates assembly. The main function of the temperature-sensitive dielectric sheath 40 is to compensate for the local electromagnetic characteristics of the power divider 30 at temperature. Specifically, the temperature-sensitive dielectric sheath 40 is made of a dielectric material with a non-zero temperature coefficient of dielectric constant, meaning that the dielectric constant of this material exhibits a predictable shift with temperature. After the temperature-sensitive dielectric sheath 40 partially covers the outer surface of the power divider 30, it can adjust the equivalent characteristic impedance and phase delay of the covered section in a manner opposite to the thermal expansion trend of the metal, thereby offsetting the impedance and electrical length shifts caused by temperature rise or fall.

[0093] The temperature-sensitive dielectric sleeve 40 is tightly fitted to the outer circumference of the power divider 30, achieving electromagnetic equivalent coupling through close contact without disrupting the metallic continuity of the power divider 30 body. The temperature-sensitive dielectric sleeve 40 does not perform electrical connection functions, but rather influences the transmission phase and equivalent impedance of the covered section through its dielectric properties.

[0094] The temperature-sensitive dielectric sleeve 40 can be fixed to the outer circular surface of the power distribution rod 30 by elastic interference fitting, local clamping, or high-frequency resistant dielectric adhesive. The slotted structure of the temperature-sensitive dielectric sleeve 40 facilitates radial opening and springback assembly, and is axially confined within a designated section to prevent axial slippage.

[0095] The material for the temperature-sensitive dielectric sheath 40 can be a polymer or composite dielectric with low dielectric loss and a controllable temperature coefficient of dielectric constant. The sheath thickness can be a thin-walled structure to avoid introducing significant additional insertion loss, while facilitating local compensation rather than full coverage.

[0096] The compensation sections and temperature sensitivity criteria are as follows: The compensation sections are several finite-length segments selected along the axial direction of the power divider 30. These segments are identified as the most sensitive to the overall standing wave or return characteristics. The temperature-sensitive dielectric sleeve 40 is only arranged on the outside of these compensation sections, rather than uniformly covering the entire length of the power divider 30. By limiting the temperature-sensitive dielectric sleeve 40 to cover only the compensation sections, temperature compensation can be applied at the most sensitive locations, avoiding the introduction of unnecessary dielectric loading in areas with high electric field strength or the main signal energy coupling path, thereby maintaining low insertion loss while controlling temperature drift. The temperature sensitivity criterion is used to determine which axial segments are classified as compensation sections. This criterion uses the upper limit frequency of the operating frequency band as a reference to evaluate the shift of the equivalent impedance and propagation phase of each axial segment with temperature changes. If the impedance or phase drift of a certain segment with temperature changes reaches a first predetermined threshold, the segment is considered a temperature-sensitive segment, and the temperature-sensitive dielectric sleeve 40 needs to be arranged.

[0097] The first predetermined threshold can be determined through thermo-electric joint simulation or hot chamber experiments. Specifically, the equivalent impedance and phase of each axial segment of the power divider 30 are extracted under different temperature conditions. These offsets are then sorted, and those segments exceeding a given stability tolerance are selected as compensation targets. This tolerance serves as the baseline for the first predetermined threshold. The first predetermined threshold can be understood as the maximum allowable local drift limit for the system to maintain standing wave or echo stability within the target operating temperature range.

[0098] The circumferential resistive film 50 is arranged on the outside of the temperature-sensitive dielectric sheath 40, and is sleeved in a basically coaxial manner. The circumferential resistive film 50 continuously covers the area in the circumferential direction, but non-conductive segmented gaps are reserved in the circumferential direction so that it is not a completely closed conductive ring; and a gradual end region is formed at both ends in the axial direction, so that the circumferential resistive film 50 gradually transitions from the effective working area to the film-free area, avoiding the formation of abrupt electromagnetic boundaries at the ends.

[0099] The circumferential resistive film 50 exhibits weak absorption characteristics with controlled surface resistance, selectively dissipating the reflected wave component returning along the axial direction while minimizing its impact on the forward wave propagating from the input to the output along the power divider 30. Due to its circumferential distribution, the film presents an equivalent energy dissipation band that is circumferentially symmetrical with respect to the coaxial field pattern, suppressing echo peaks and intracavity ringing.

[0100] The circumferential resistive film 50 and the temperature-sensitive dielectric sheath 40 work together in synergy. The circumferential resistive film 50 is located outside the temperature-sensitive dielectric sheath 40, forming a double-layer structure. The temperature-sensitive dielectric sheath 40 mainly provides temperature compensation, while the circumferential resistive film 50 mainly provides echo dissipation. When the two are superimposed, temperature drift suppression and reflection peak passivation can be achieved simultaneously in the same physical segment without the need to add an additional independent mechanical module to the power divider 30.

[0101] The circumferential resistive film 50 can be formed on the outer surface of the temperature-sensitive dielectric sheath 40 by means of attachment, coating, deposition, printing, or transfer bonding. The film layer must be kept in close contact with the temperature-sensitive dielectric sheath 40 to avoid the formation of cavity-like gaps under high-frequency fields.

[0102] The circumferential resistive film 50 can be a conductive polymer film with controllable sheet resistance, a carbon-based composite conductive layer, a composite coating containing conductive fillers, etc. The film layer as a whole is strip-shaped or ring-shaped, and can be one or more discrete ring segments along the axial direction.

[0103] It should be noted that the circumferential resistive film 50 is not applied to any arbitrary location on the power divider 30, but only to the outer side of the power divider 30 at the location determined to be the low-field ring zone. The low-field ring zone refers to those annular regions at the upper limit frequency of the operating frequency band where the normalized local electric field amplitude of the preceding wave component is lower than a fourth predetermined upper limit ratio; it can be understood as the relative nodal region of the electric field amplitude. Strictly confining the circumferential resistive film 50 to the low-field ring zone means that the circumferential resistive film 50 mainly couples with the dwell energy of the return wave, without significantly absorbing the energy of the main transmitted preceding wave. This allows for a concentrated reduction in the return peak value without significantly increasing insertion loss.

[0104] The fourth predetermined upper limit ratio is determined by normalizing the circumferential electric field distribution at each axial position outside the power divider 30 through field simulation or field measurement. Using the main propagation component of the advancing wave as a reference, regions with weaker electric fields are defined as nodal regions. If the local electric field amplitude within a certain ring is lower than the normalized specific ratio threshold, the ring is considered to have a relatively small impact on the main propagation wave energy and can support the circumferential resistive film 50. This specific ratio serves as the benchmark for the fourth predetermined upper limit ratio. This ratio can be set by comparing the insertion loss degradation and return peak suppression at different film application positions, selecting the boundary point where the insertion loss remains at an acceptable level while the return suppression effect is significant as the fourth predetermined upper limit ratio.

[0105] It should be noted that the circumferential resistive film 50 has non-conductive segmented gaps in the circumferential direction. That is, the circumferential resistive film 50 does not form a completely closed annular conductive band, but is broken into several segments, with electrical isolation gaps between each segment. Simultaneously, the circumferential resistive film 50 has a gradual transition at both ends in the axial direction, causing the film thickness, conductivity, or coverage width to gradually transition to zero. The segmented gaps prevent the formation of closed loops, preventing unexpected circulating currents or resonant circuits in high-frequency fields; the gradual transition avoids sharp electromagnetic boundary changes at the film's termination position, reducing localized reflections or electric field concentration at the ends. The segmented gaps and the gradual transition together ensure that the circumferential resistive film 50 acts as a gentle, fixed-point damping surface, rather than a new discrete reflecting surface.

[0106] Under typical operating conditions, the radio frequency signal is coupled into the power divider 30 through the input port 210 and distributed to multiple output ports 220. The metal structure of the power divider 30 undergoes thermal expansion or material parameter drift under high-frequency and temperature-dependent conditions, making the equivalent impedance and phase delay of certain axial sections extremely sensitive to temperature; these sections are the compensation sections. The temperature-sensitive dielectric sleeve 40, according to temperature sensitivity criteria, is installed only on the outside of these compensation sections. The direction and magnitude of the dielectric constant change of the temperature-sensitive dielectric sleeve 40 with temperature cancel out the trend of changes in the metal geometry and electrical length of the power divider 30, thereby locally correcting the impedance and phase of this section and reducing the impact of temperature drift on the overall standing wave and echo.

[0107] Meanwhile, some energy will be reflected back and forth between the power divider 30 and the cavity 10, forming echo peaks or hot spots where resident energy is concentrated, and may affect the isolation between the output ports 220 through coupling paths. The circumferential resistive film 50 is only applied to the low-field ring zone selected by the node criterion, that is, the location where the electric field of the advancing wave is lower and the field components of the reflected wave or stagnant wave are more significant. The circumferential resistive film 50 forms a controlled dissipation band at these locations, dissipating the energy of the returning wave, weakening the echo peak, reducing intracavity ringing, and at the same time almost not absorbing the main transmission advancing wave, thus having a very low impact on insertion loss.

[0108] The segmented gaps prevent the circumferential resistive film 50 from forming a continuous closed loop and introducing new coupling paths, while the gradual termination avoids sharp field abrupt changes at the termination position of the circumferential resistive film 50. Therefore, this double-layer structure simultaneously suppresses full-band shifts caused by temperature drift and narrow-band spikes caused by local reflection peaks during operation.

[0109] Under normal steady-state operation, the temperature-sensitive dielectric sheath 40 adheres stably and changes in temperature in tandem with the power divider rod 30. Because the dielectric constant of the temperature-sensitive dielectric sheath 40 has a predictable temperature coefficient, it dynamically fine-tunes the equivalent electrical length and impedance of the compensation section, keeping the standing wave nearly constant at different temperatures, rather than drifting significantly with ambient temperature.

[0110] If a significant phase shift occurs in an uncompensated section during thermal shock or a rapid rise in ambient temperature, the section will be included in the compensation section and covered with a temperature-sensitive dielectric sleeve 40. The temperature-sensitive dielectric sleeve 40 will immediately generate a reverse correction effect, thereby preventing echo degradation of the entire machine under high or low temperature conditions.

[0111] When there is port mismatch or an undesirable external load, the return wave travels along the power divider 30 and forms a relatively stable local field distribution at a specific axial position. The circumferential resistive film 50 is positioned specifically to address these energy retention points of the return wave, causing it to attenuate the energy through heat dissipation, thus reducing inter-port crosstalk and spike echoes. Since the circumferential resistive film 50 is only deployed in the annular region where the electric field amplitude is below a fourth predetermined upper limit, the main transmission component is hardly dissipated. Therefore, the main signal path of the output port 220 can still maintain low additional loss under different load conditions.

[0112] After long-term operation or multiple thermal cycles, the segmented fracture can suppress the potential stress accumulation of the film layer as an integral conductive ring, and the gradual termination can mitigate the risk of peeling at the film layer ends, thereby improving the mechanical stability and service life of the double-layer structure.

[0113] The determination of the compensation section can rely on simulation analysis, or it can be achieved by heating or cooling the prototype point by point under constant input conditions to scan the standing wave changes of each axial segment, thereby identifying the most sensitive segment. The node criteria can also be obtained through simulation of the field distribution, or by measuring the impact of local application on insertion loss and echo using a small-area removable loss patch trial application method on the prototype, finding the position that can suppress the echo peak value to the greatest extent without significantly increasing the insertion loss, and then marking this position as the low-field ring zone.

[0114] In a further integrated scheme, the temperature-sensitive dielectric sheath 40 and the circumferential resistive film 50 not only function independently but also synergize with the aforementioned other structures: the segmented conical profile with multiple slopes and the segmented design with equal field energy allow the field energy distribution and reflection phase distribution of the power dividing rod 30 along the axial direction to be artificially shaped, and the positions of the nodes and anti-nodes can be predicted in advance; based on this known distribution, the circumferential resistive film 50 can be attached only to the node ring, achieving echo suppression with almost no increase in insertion loss. The mode suppression structure on the inner wall of the cavity 10 and the impedance trench 140 at the port jointly improve the equivalent cutoff of higher-order modes and establish an approximate virtual ground at the port surface, raising the threshold of the higher-order mode coupling path, while the circumferential resistive film 50 attached to the node ring further absorbs the remaining excitable higher-order modes and the energy of the back and forth echoes, and the two work together to significantly weaken the ringing in the cavity. The multi-slope segmented conical profile and microstructure 320 disperse and homogenize the reflection phase of each segment on the body. The temperature-sensitive dielectric sheath 40 further cancels out the small phase and equivalent impedance drift caused by temperature, thus maintaining low standing wave and low echo over a wide temperature range. This synergy also provides a buffer against manufacturing and assembly tolerances: the body structure reduces tolerance sensitivity through geometric gradients, and the temperature-sensitive dielectric sheath 40 and the circumferential resistive film 50 achieve self-correction during operation through later application, ensuring high consistency of the whole under batch assembly, environmental changes, and aging conditions.

[0115] The objectives of the above-mentioned synergy can be summarized as follows: the effective bandwidth corresponding to the available echo index range is significantly improved, the port VSWR peak value is further reduced, the port isolation is improved, the insertion loss degradation is controlled within a very small range, and the VSWR curve is kept stable within a significant temperature change range, so that the stability of the whole machine under high temperature or low temperature conditions is close to that under normal temperature conditions.

[0116] In this embodiment, by employing the technical means of setting a temperature-sensitive dielectric sleeve 40 on the outer surface of the power divider 30 and arranging the temperature-sensitive dielectric sleeve 40 only in the compensation section based on the temperature sensitivity criterion, and setting a circumferential resistive film 50 on the outside of the temperature-sensitive dielectric sleeve 40 and applying it only in the low-field ring based on the node criterion, and having segmented seams along the circumferential direction and a gradually tapered end at the axial end edge, the technical problems of traditional power dividers, such as the deterioration of standing wave ratio caused by impedance and phase drift with temperature changes when the ambient temperature changes, and the high echo peak value and difficulty in reducing inter-port crosstalk when port mismatch and cavity reflection accumulation occur, are effectively solved. Thus, the technical effect of maintaining low standing wave ratio, low echo, high isolation and low insertion loss and improving long-term stability is achieved simultaneously under wide bandwidth and wide environmental conditions.

[0117] The following is an implementation example of a power divider based on the above multi-structure collaborative design. The power divider rod 30 is a monotonic, multi-slope segmented cone; each segment has a circumferential periodic microstructure 320 on its outer circle with a clear distance between segments; a dense area of ​​microstructures 320 is arranged in the high-slope segment; an annular anti-modulation groove is placed on the inner wall of the cavity 10 at this position; a λ / 4 impedance groove 140 is made on the output end face; the segment length is determined according to the principle of equal field energy; and a temperature-sensitive dielectric thin sleeve 40 and an annular resistive film 50 are added in the temperature-sensitive segment. The overall goal is to achieve flatter standing wave and echo peaks and valleys, higher isolation, and better passivation of assembly / temperature tolerances. The specific details of this power divider are as follows:

[0118] I. Design Conditions: Operating Frequency Band: 0.70GHz~2.70GHz. 2.70GHz is defined as the upper limit frequency fhigh of the operating frequency band. (The last part, "f," appears to be a typo and can be omitted.) high The field distribution, waveguide wavelength, and skin depth serve as the "upper limit frequency references" for all frequencies. Dielectric space: Approximately air, with an equivalent relative permittivity εr≈1 (no overall dielectric block is filled; local support / compensation structures are only allowed at specified locations). Cavity 10 (outer conductor): Cavity 10 is a highly conductive metal shell (copper alloy / silver-plated aluminum alloy, etc.), axially continuous, forming the first end 110 (input) and the first end 120 (output). The inner wall of cavity 10 is a continuous conductive surface (coaxial outer conductor), with a reference inner radius b≈14.0mm (i.e., shell inner diameter ≈28mm). Externally, it features conventional RF cavity 10 characteristics such as positioning shoulders, threaded / flange fixing surfaces, and sealing grooves. Input port 210: N-type or equivalent 50Ω coaxial interface, located at the first end 110, with the port axis coaxial with the cavity 10 axis. Output port 220: Two equal-power outputs, located at the first end 120, with the two output ports symmetrically distributed on the same end face. Each output terminal corresponds to one output branch on the power divider rod 30 (1-input, 2-output architecture). Power divider specifications (targets) include: Input port 210 echo (|S11| converted VSWR): return loss RL ≥ 18dB full band; single-path insertion loss IL ≤ 0.30dB (additional losses on the basis of ideal 3dB power divider should be minimized); isolation between the two output ports 220 ≥ 25dB full band; minimal in-band echo peak and valley fluctuations (suppressing the phenomenon of "narrowband spikes + band edge ripples"); stable VSWR curve over a wide temperature range (e.g., -40℃ to +85℃) without obvious drift steps; high robustness to assembly tolerances and repeated assembly and disassembly; tolerance for slight port mismatch without immediately turning cavity 10 into a high-order mode resonant cavity. Total axial space: an effective "modulation length" L ≈ 50mm is usable from the input coaxial surface to the output flange reference surface; cavity 10 should not be significantly lengthened (traditional modification methods often lengthen the transition section to suppress VSWR).

[0119] II. The overall structural framework of the power divider is as follows:

[0120] The cavity 10 is a single piece of metal housing that runs through the entire shaft. The first end 110 has an input coaxial port, and the first end 120 has two output coaxial ports. The input inner conductor is electrically connected to the front end of the power divider 30; the two output inner conductors are respectively connected to the two output branch ends of the power divider 30.

[0121] The power divider 30 (inner conductor) is located on the central axis of cavity 10 and is the core of the entire gradient / distribution process. It is not a simple step or a single cone, but a "multi-slope segmented conical profile that monotonically changes along the axial direction": the outer diameter increases monotonically from a small radius near the input end until it reaches a larger radius near the output branch area; it is divided into multiple impedance transformation segments 310 (e.g., 12 segments) along the way, with smooth transitions between adjacent segments using rounded corners, rather than abrupt steps; "multiple slopes" means that these 12 segments are further divided into several macroscopic regions: a slow slope in the first segment, a relatively fast slope in the middle segment, and a slow slope in the last segment, forming a monotonous outer diameter curve with an "S"-shaped rhythm; this monotonous gradient of "slow-fast-slow", combined with the segment length distribution, can disperse the phase of small reflections at different frequencies and spread them out evenly, rather than having them all in phase and superimposed at a certain frequency. On the outer circumferential sidewall of each impedance transformation section 310 of the power divider 30, microstructures 320 (circumferential shallow grooves 321 / shallow corrugations / rounded micro-tooths) are arranged periodically along the circumferential direction; these microstructures 320 do not cover the entire axial length of the entire section, but only cover the middle area of ​​60% to 80% of the axial length of the section.

[0122] Near each segment boundary, a smooth transition zone, known as the "first preset spacing," must be left to prevent scattering from the segment boundary itself and scattering from the micro-texture from overlapping at the same location; the undulations (groove depth) of the microstructure 320 are limited to not exceeding the material of the power divider 30 at f highA certain proportion of the skin depth δ (typically no more than about 20%·δ, where δ can be roughly estimated at ~1.3μm for high-conductivity copper in the 2.7GHz range; that is, the groove depth is within 'several times the sub-skin depth', which is a mild disturbance to the current path, rather than cutting out an independent resonant cavity). In the middle section of the power divider 30 (i.e., the region with the largest geometric slope), an axial length is selected as the "microstructure 320 dense region": here the circumferential shallow grooves 321 have a shorter pitch and a denser number (high-density microstructure 320); the purpose is to further refine and disperse the phase of these inevitable small reflections in the section where the geometric changes are most drastic and local large reflections are most likely to occur; this dense region also maintains a first preset distance from the adjacent section boundary, and the transition rounded corners should be softer to avoid forming high-field peaks; the inner wall of the cavity 10 (inner surface of the outer conductor) is provided with a circumferential suppression recess at a position opposite to the "microstructure 320 dense region" axially. Slot array (impedance mode suppression structure 130); the mode suppression grooves are 360° closed annular shallow grooves opened along the inner wall of cavity 10, arranged in rows along the axial direction; the groove depth, groove width, and groove spacing of each groove are limited to only a small fraction of the wavelength of the TEM main mode guided wave at fhigh (typically each term is much smaller than λhigh / 4, which can be conventionally understood as "electrically a fine-scale disturbance"); they do not strongly reflect the main mode, but raise the equivalent cutoff threshold of higher-order modes (TE11, TM01, etc.) near this cross section to prevent the "small reflections scattered on the power divider 30 side" from running along the cavity wall into residing higher-order mode ringing.

[0123] On the first end face 120 where output port 220 is located, there is an open-loop "impedance trench 140" (that is, the λ / 4 virtual ground trench on the port end face) surrounding each output coaxial interface: this trench is designed according to the electrical length at f high The electrical length of the guide wave is close to one-quarter of the TEM master mode, making the area near the port plane appear as an approximate virtual ground (high suppression boundary); this can significantly improve the isolation between the two output ports 220 and suppress the path of "current conduction from the port flange to another path"; the impedance trench 140 and the mode suppression groove of the inner cavity maintain a second preset distance in the axial direction, avoiding the two modulation structures from being directly coupled into a new narrowband resonant cavity.

[0124] The lengths of the segments mentioned above are not arbitrarily cut, but determined according to the principle of "segmentation based on equal field energy": in f high At the most demanding frequency, the annular medium space between the outer surface of the power divider 30 and the inner wall of the cavity 10 is regarded as the main energy channel; for each candidate segment, the electromagnetic energy density (or equivalent electric / magnetic field energy) within that segment is integrated; the axial length of the segment is adjusted so that the energy integral values ​​borne by adjacent segments are as close as possible, and the difference between the two is less than the larger one by a set threshold (e.g., 15%); reflections will not accumulate in a single place, and echo peaks will not emerge at the boundary of a single segment.

[0125] Regarding temperature drift and high-power steady-state conditions: For axial sections (defined as compensation sections) where thermal simulations / experiments have shown that "the equivalent impedance or phase of this section drifts significantly as the temperature rises," a slit cylindrical temperature-sensitive dielectric sleeve 40 is locally fitted around the outer circumference of the power divider 30. The temperature-sensitive dielectric sleeve 40 is made of a low-loss material with a controllable dielectric constant that changes with temperature (the temperature coefficient of the dielectric constant is non-zero). It effectively produces a change in impedance and phase of this section opposite to the direction of thermal expansion of the metal. Correction is made to offset temperature drift; on the outside of the temperature-sensitive dielectric thin sleeve 40, a ring of circumferential resistive film 50 is wrapped only in the electric field node region (where the field strength after normalization of the advancing wave is lower than a certain percentage threshold, such as <20% peak value): this film is segmented (with gaps, not forming a complete closed conductor), and the two ends of the axis are gradually tapered; the ring of resistive film 50 slightly absorbs the round-trip reflected / residual energy, but hardly absorbs the main advancing wave, so it can suppress the return wave spike and suppress ringing, without significantly increasing the insertion loss.

[0126] III. Axial geometry of power divider 30 (inner conductor): Multi-slope monotonic piecewise cone and equal-energy piecewise section, as detailed below:

[0127] Basic aperture and impedance direction: Assume the input port 210 is 50Ω, and the radius of the coaxial outer conductor is b≈14.0mm. The corresponding radius a of the power divider rod 30 near the input terminal is... in It is on the order of 6.1 mm (this is just an illustration; the actual value can be deduced using the standard 50Ω coaxial approximation).

[0128] The goal is to gradually pull the 50Ω impedance towards the internal equivalent combined impedance (two-way equal power output equivalent to 25Ω level), so the radius of the power divider rod 30 increases monotonically along the axial direction to a. out Approximately 9.2 mm.

[0129] The total axial length can be L≈50mm, within which the entire gradient and power distribution transition from 50Ω to approximately 25Ω is completed.

[0130] Multi-slope zones: Zone 1 (front section, approximately 0-27% of total length, 0-13.5mm): Gentle slope. Purpose: To reduce the overall reflection noise at the low end of the bandwidth, avoiding a sudden change at the beginning. Zone 2 (middle section, approximately 27%-73% of total length, 13.5-36.5mm): Larger slope. Purpose: To quickly push the impedance to the target range in the middle section; local small reflections are most likely to occur here → at the same time, a "micro-structure 320 dense area" will be arranged here. Zone 3 (final section, approximately 73%-100% of total length, 36.5-50mm): Slows down the slope again. Purpose: To slow down near the output branch for fine-tuning, avoiding strong standing wave spikes near the output end.

[0131] Segment Number Example (N=12 segments): The following lists 12 segment schematics, each segment being an impedance transformation segment 310. z-start / end: ​​The axial coordinates of the start / end point of this segment (mm, measured from the input end face z=0); L i : Axial physical length of this segment (mm); a start to end: Monotonic change of radius of the power dividing rod 30 from the start to the end point (mm); Microstructure 320 pitch p i : Repeating pitch (mm) of the circumferential shallow groove / micro-tooth in the axial direction; groove depth h i Typical radial undulation (mm) of a single shallow groove, constrained by "groove depth ≤ upper limit of skin depth ratio"; Coverage ratio: the axial ratio covered by the microstructure 320 in this segment (the remainder is reserved as a smooth transition area for the first preset spacing).

[0132]

[0133] Key points: a. Start to End: The entire sequence only increases and never decreases, strictly monotonous, avoiding reverse inflection points, as these are most prone to phase overlap leading to large spike echoes. All segment connections should use fillet radii r≈0.6~0.8mm (can be adjusted according to f). high 1% to 2%·λ high (Upon arrival at the school), to avoid sharp steps causing edge field spikes. Each microstructure 320 does not touch the segment boundary: a smooth 0.8mm is left at the beginning and a smooth 0.8mm is left at the end of the segment; this is the physical implementation of the "first preset spacing". Zone 2 (segments 4 to 7, axial distance approximately 13.5 to 28.8mm) is defined as the "microstructure 320 dense zone": p i Smaller (0.45mm level, compared to 0.75mm level in Zone 1) and with higher coverage (80% axial coverage), this effectively enhances "phase dispersion and edge field passivation" in the areas most prone to localized small reflections. These values ​​(p i h i Both coverage ratio and coverage must meet two hard constraints:

[0134] (1) Trench depth h i It must be below a "first predetermined ratio" based on the skin depth δ (e.g., no more than 0.2 to 0.3 times δ). In this way, the shallow groove only "gently disturbs the surface current path" and does not become a separate resonant cavity or a huge source of additional losses.

[0135] (2) Axial pitch p i Must be less than f high Corresponding waveguide wavelength (TEM dominant mode waveguide wavelength λ) high To avoid forming a significant Bragg cycle, a certain "small fractional upper limit" is set; typically, the pitch is controlled at λ. high Less than 5% to 10%. (f) high=2.7GHz, λ≈111mm in air, therefore the pitch of 0.45mm~0.75mm is much smaller than 10% of 111mm, which is electrically a fine texture and will not open a "bandgap" on its own.

[0136] IV. Synergy between the "320 dense microstructure area" and the mold-suppressing groove on the inner wall of cavity 10:

[0137] The "microstructure 320 dense area" refers to the area from segment 4 to segment 7 (z≈13.5mm~28.8mm). There, the radius of the power divider 30 increases the most, the impedance is pushed significantly towards the target value over a short distance, and small reflections are most likely to occur.

[0138] In the opposing axial region of this area, the inner wall of cavity 10 has several circumferential mold-suppressing grooves (impedance mold-suppressing structure 130):

[0139] For example: Place two 360° annular slots at z≈18mm and z≈30mm respectively; the slot depth d≈1.1mm, the slot width w≈2.2mm, and the slot spacing (center-to-center distance between two slots) ≈12mm; all slot openings are rounded (R≥0.3mm) to avoid sharp edges and current accumulation; these dimensions must be constrained by a set of "upper limit fractions": the slot depth / slot width / slot spacing must not exceed the limits set by the TEM master mode at f high waveguide wavelength λ high The upper limit of the small fractions (e.g., each not exceeding 0.02 to 0.15·λ) high (On this order of magnitude). The goal is to make it exhibit gentle boundary coarsening for the main mode, while acting as a "threshold" for higher-order modes (TE11, TM01, etc.), raising the equivalent cutoff frequency of higher-order modes and preventing them from easily dwelling and ringing.

[0140] Key point: The mold-suppressing groove on the inner wall of cavity 10 and the "micro-structure 320 dense area" of power distribution rod 30 are axially opposed.

[0141] The 320-degree microstructure dense region breaks down local small reflections into more and finer small reflections and introduces more near-field components from the walls. The opposite section of the cavity wall uses circumferential mode-suppressing grooves to directly "block" the propagation path of these high-order modes of near-wall energy, preventing them from becoming high-order mode resonants within the cavity. These two structures, one above the other and one inside the cavity, together effectively "disperse and intercept" the flow.

[0142] 5. Output port 220 end face impedance trench 140 (λ / 4 virtual ground ring groove) and second preset distance.

[0143] On the first end 120, which is the end face where the two output ports 220 are located, one or two annular grooves (impedance grooves 140) are circumferentially formed in the outer conductor flange area of ​​each output port. This is the structure you mentioned in your text that "approximately virtual ground boundaries are established on the port faces to suppress coupling between ports".

[0144] The geometric depth, opening width, and cross-sectional shape of the impedance trench 140 (which can be a single annular trench or a double-annular trench with a shallow outer edge and a deep inner edge) are not simply a matter of "the deeper the better," but rather the goal is to ensure that its equivalent electrical length is within f. high It is approximately one-quarter of the wavelength of the dominant mode. high At 2.7 GHz, the quarter wavelength in free space is approximately 111 mm / 4 ≈ 27.8 mm. Actual trenches are not purely air gaps; they have metal walls, concentrated electric fields, and opening capacitance, so the physical depth will be shorter than 27.8 mm. Empirically, a "two-stage trench" can be constructed: an outer layer slightly shallower (e.g., 10–12 mm deep, 3.5 mm wide opening), and an inner layer slightly deeper (e.g., 18–22 mm deep, 4.0 mm wide opening), with continuous layers but rounded corners. Then, electromagnetic simulation is used to extract the equivalent electrical length and scale it by a factor (similar to 0.85–0.95) to bring it close to λ. high The equivalent point is / 4". This impedance groove 140 section must maintain a "second preset distance" in the axial direction with the molding groove on the inner wall of the aforementioned cavity 10, typically allowing a net distance on the order of >5mm to >10mm (approximately equal to λ). high (Level 0.05).

[0145] Objective: To ensure that the "mode suppression section within cavity 10" (responsible for blocking higher-order modes and suppressing ringing within the cavity) and the "virtual ground trench on the port surface" (responsible for improving isolation and blocking bypass coupling between ports) function independently, preventing them from combining to form a new narrowband cavity. When a significant mismatch occurs at one of the output ports 220 (due to an undesirable external load), the impedance trench 140 will locally contain the mismatch energy near that port, instead of allowing it to propagate along the end face metal surface and the wall of cavity 10 to another port, thus maintaining good inter-port isolation even under abnormal operating conditions.

[0146] 6. Implementation of the "equal field energy segmentation" difference threshold (second predetermined upper limit ratio).

[0147] The 12 segments mentioned above are not of equal length; segments 4 to 9 are significantly shorter because the middle segments have a steeper slope and are more energetic, requiring finer segmentation. The practical approach is: using f... highA three-dimensional field solution (electric field, magnetic field) is performed. The energy density is integrated axially over the annular medium space (between the outer circle of the power divider 30 and the inner wall of the cavity 10) of each candidate segment to obtain the "energy integral value" for each segment. A second predetermined upper limit ratio is set, for example, 15%. For any pair of adjacent segments, the difference in their energy integrals is divided by the larger of the two; the result must be ≤15%. If this is not satisfied, the segment length is redistributed between the two segments to achieve a more balanced energy distribution. After equalization, a single segment will not bear an overwhelming amount of field energy, thus avoiding particularly sharp reflection hotspots at the segment boundaries. The in-band echo curve therefore becomes flatter, rather than "a sudden 28dB peak at a point of 2.2GHz." In the above example of 12 segments, it can be understood that segments 4 to 9 are energy-dense segments, which are cut short and densely packed with microstructures 320, directly opposite the mode-suppressing groove on the inner wall of the cavity 10. This is precisely the region where "energy equalization + reflection reduction + interruption of higher-order mode coupling" are superimposed simultaneously.

[0148] 7. Temperature-sensitive dielectric sheath 40 and circumferential resistive film 50 (temperature drift compensation and echo dissipation).

[0149] This pair is a post-processing / post-assembly pluggable "correction layer" placed on a specific axial window on the outer circle of the power divider 30, stacked outside the body metal.

[0150] 7.1 Temperature-sensitive dielectric thin sleeve 40:

[0151] Structural Form: A thin-walled, slotted cylindrical dielectric sleeve is partially fitted onto the outer circumference of the power divider 30. "Slotted" means a narrow longitudinal slit is left along the circumference, preventing the formation of a completely closed loop and avoiding the formation of a closed conductive ring or strong cavity; it also facilitates light pressure insertion and provides slight elastic clamping. The temperature-sensitive dielectric sleeve 40 is made of a low-loss dielectric material, and its dielectric constant exhibits a predictable linear / quasi-linear drift with temperature (the temperature coefficient of the dielectric constant is not zero). The goal is that when the power divider 30 is heated, the thermal expansion of the metal will drag the local equivalent impedance and phase away from the nominal value; the direction of the dielectric parameter change of the temperature-sensitive dielectric sleeve 40 with temperature is deliberately reversed, pulling the overall equivalent impedance / electrical length back towards the nominal value. It is only installed in the "compensation section"—the so-called compensation section refers to those temperature-sensitive areas determined through thermo-electrical joint analysis; the definition standard is: within f... high If the equivalent impedance and / or phase shift of a certain segment with temperature exceeds a first predetermined threshold during the test, that segment is designated as a compensation segment. The temperature-sensitive dielectric sheath 40 is not required to cover the entire power divider rod 30; it only appears in a few compensation segments (e.g., a few millimeters between segments 5 and 6, or near segment 10), and its length is limited to those few millimeters; thus, it almost never touches the segment head / boundary where the electric field peak is highest, nor does it increase the dielectric loading loss of the entire segment.

[0152] 7.2 Circumferential resistive film 50:

[0153] On the outer surface of the temperature-sensitive dielectric sheath 40, a ring of circumferential resistive film 50 is attached. This film is a high surface resistivity, low thickness conductive composite layer (carbon-based / conductive polymer, etc.), which wraps around the outside of the temperature-sensitive dielectric sheath 40 in a "ring". However, it is not a complete closed ring: it is broken into several segments along the circumference, with narrow slits in the middle to avoid forming a complete current loop; at both ends of the axial direction, the film layer gradually tapers off (gradual edge reduction), so that the presence of the film slowly changes from "present" to "absent", rather than an abrupt step. This gradual edge reduction can be achieved by gradually reducing the film thickness, gradually reducing the coverage width, or gradually reducing the concentration of conductive filler. The circumferential resistive film 50 is only allowed to be attached to the "low field ring zone", that is, the axial positions (node ​​regions) around the outer circle of the power divider 30 where the electric field amplitude is the lowest: node criterion: under f_high, normalize the advancing wave component and find those circumferential zone regions where the local electric field amplitude does not exceed the fourth predetermined upper limit ratio (e.g., 20% or less); the main energy of the advancing wave is very low in these places; placing the circumferential resistive film 50 here means that it mainly couples the round-trip reflection / dwelling energy, rather than the main advancing wave, so it basically does not increase the insertion loss, but can "eat" the reflection spikes and small ring rings.

[0154] The "thermal dielectric sheath 40 and circumferential resistive film 50" achieve two functions: anti-temperature drift: the temperature coefficient of the dielectric constant of the thermal dielectric sheath 40 cancels the thermal drift of the metal in the compensation section, and the standing wave curve does not drift by a large increment between -40℃ and +85℃; anti-local echo spikes: the circumferential resistive film 50 dissipates the energy of the round-trip reflection at the node, which is equivalent to a fixed-point damper, suppressing the narrowband spike of |S11|, shortening the ringing life in the cavity, and reducing the chance of crosstalk between ports.

[0155] Mechanically: The temperature-sensitive dielectric thin sleeve 40 uses its own elasticity to clamp the outer circle of the power distribution rod 30 without welding it, which is convenient for debugging and replacement; the circumferential resistive film 50 is attached to the outer surface of the temperature-sensitive dielectric thin sleeve 40 without directly scraping the metal of the power distribution rod 30, and can be replaced later or different resistive levels can be selected; the segmented breaks and gradual ending can also effectively alleviate thermal stress and prevent the entire film from breaking after long-term stretching.

[0156] VIII. Key Points of the Debugging Process

[0157] First-round verification process: Full-band S-parameter scan at room temperature: Observe the input echo |S11|, and it is required that there should be no high and sharp isolated peaks in the peak-to-valley difference within the band; Observe the flatness of the insertion loss |S21| and |S31| of the two outputs; Observe the isolation |S23| between the two outputs, which should be ≥25dB ​​across the full band; Extract the group delay to ensure that there is no abnormal narrowband resonance in the middle section (320 dense area of ​​microstructure).

[0158] Modal energy projection analysis: Near the mid-section and output end, observe the energy ratio of higher-order modes such as TE11 / TM01. It should be suppressed by the combination of mode suppression groove and impedance trench 140.

[0159] Temperature scanning: For example, -40℃, 25℃, +85℃, repeat the frequency sweep; the standing wave and isolation curves should maintain the same "shape", only allowing small overall shifts or stretches, rather than sudden new peaks appearing in local areas.  If the echo suddenly deteriorates in a certain frequency band at a certain temperature point, check whether a temperature-sensitive dielectric thin sleeve of 40mm can be added / adjusted in the corresponding axial section.

[0160] Port mismatch / high power tolerance: Intentionally load one of the output ports 220 with a slightly mismatched load, then scan the echo and isolation of the other port; observe whether the mode suppression groove and the λ / 4 impedance groove 140 can still maintain high isolation and low standing wave ratio, and prevent energy from ringing in the cavity 10 in an "echo chamber" style.

[0161] PIM / High-power stability: Use the dual-tone test method to check intermodulation; if there is any abnormality, check the cleanliness of the end face contact, the integrity of the coating, and whether the flange tightening torque is uniform.

[0162] IX. In conclusion:

[0163] Monotonic, multi-slope segmented conical power divider 30: The outer diameter increases monotonically, and the impedance changes gradually in one direction, leaving no opportunity for a reverse inflection point and avoiding local phase convergence. The three-slope "gradual-rapid-gradual" design suppresses low-frequency noise, rapidly pushes up the mid-section impedance, and then slows down at the tail end to stabilize the matching near the output port. Adjacent segments are connected with rounded corners to suppress edge field peaks and additional losses.

[0164] Circumferential periodic microstructure 320 and dense region of microstructure 320: shallow circumferential grooves 321 (rounded micro-tooth) are made on each segment of the outer circumference, and the axial period is much smaller than λ. high The grooves are deeper than the upper limit of the skin depth, belonging to "controlled perturbations". These perturbations break the unavoidable small reflections within the segment into many small reflections, and the phases are dispersed. The most severe segments (region 2) are densified to form "dense areas", specifically to deal with the places most prone to echo spikes. The shallow grooves of each segment maintain a first preset distance from the segment boundary to avoid the scattering of the shallow grooves superimposed on the scattering that already exists at the segment boundary.

[0165] Mode suppression grooves (impedance mode suppression structure 130) on the inner wall of cavity 10: An array of annular grooves 130 is arranged axially on the cavity wall directly opposite the dense area; the geometry of the grooves is constrained by the "upper limit fraction", ensuring that they only "coarse the boundary" electrically and raise the equivalent cutoff of higher-order modes, without attacking the main mode. This area is equivalent to a coaxial double layer of "reflection refinement and mode suppression gate": the power divider 30 side refines the reflection into fragments, and the cavity wall side blocks these fragments from becoming higher-order mode standing waves along the wall surface.

[0166] Output port 220λ / 4 impedance trench 140: An annular trench (can be double-stage) is dug on the port flange surface, adjusted to approximately one-quarter of the waveguide wavelength according to the equivalent electrical length, creating an approximate virtual ground boundary at the port opening. This significantly improves isolation, preventing crosstalk between ports via housing surface currents, especially in cases of mismatch or high power. This trench and the mode-suppressing groove maintain a second preset axial distance to prevent them from combining to form a small resonant cavity.

[0167] Equal-energy segmentation / second predetermined upper limit ratio: The segment length is not mechanically averaged, but rather "equal-energy segmentation" is performed according to the field distribution of f_high: the energy integral difference between adjacent segments is controlled within a certain upper limit ratio (e.g., 15%). As a result, small reflection contributions are evenly distributed across all segment boundaries, rather than exploding into large peaks at a single segment boundary. This directly flattens the in-band echo peaks and valleys, widening the usable bandwidth.

[0168] Thermosensitive dielectric sleeve 40 and circumferential resistive film 50: A thin dielectric sleeve with a temperature coefficient of dielectric constant is sleeved on the most heat-sensitive local section to counteract the impedance / phase drift caused by metal thermal drift, fixing the standing wave curve within the range of -40℃ to +85℃. Outside this thermosensitive dielectric sleeve 40, a ring of circumferential resistive film 50 with a broken seam and a gradually tapering end is attached. Its position is chosen at the electric field node zone, so that it hardly consumes the main advancing wave power, but can gently dissipate the round-trip reflected energy, suppress narrowband echo spikes, shorten the ringing life within the cavity, and suppress crosstalk. This is the back-end calibration structure of "temperature drift self-compensation and fixed-point damping," which also allows for batch assembly / reassembly: only the sensitive section is covered, eliminating the need for rework of the entire machine.

[0169] In summary, this is a broadband coaxial power divider solution with a total length L≈50mm and the same two-way power divider configuration, but with wider bandwidth, flatter standing wave ratio, higher isolation, and less sensitivity to temperature and assembly. The reflected phase of this power divider is no longer concentrated at a single point; the segment boundary edge field is passivated by rounded corners and a 320° microstructure; higher-order modes on the wall are blocked by mode-suppression grooves; the port plane uses λ / 4 grooves to create an approximate virtual ground, cutting off crosstalk between ports; the segment length is averaged to bear the reflection contribution through the equal field energy criterion; the temperature-sensitive dielectric sheath 40 and the circumferential resistive film 50 provide "secondary insurance" for temperature drift and residual standing wave ratio in the later stages.

[0170] The above description is merely illustrative of the invention. Those skilled in the art can make various modifications or additions to the described specific embodiments or use similar methods to replace them, as long as they do not depart from the content of this specification or exceed the scope defined by the claims, all of which should fall within the protection scope of this invention.

Claims

1. A power divider, comprising: The cavity has a first end and a second end that extend through it along its length. A port assembly includes an input port and multiple output ports. The input port is located at the first end and is coaxially arranged with the cavity axis. The multiple output ports are all located at the second end. The power divider is coaxially disposed inside the cavity and has an input end electrically connected to the input port and multiple output branch ends electrically connected to the multiple output ports respectively. The input end is electrically connected to the input port, and each output port is connected to each output branch end in a one-to-one correspondence. Its features are: The power divider is composed of multiple impedance transformation segments connected sequentially along the axial direction. The outer diameter of the power divider from the input end to the output branch end varies monotonically along the axial direction and forms a segmented conical profile with multiple slopes. Microstructures are periodically distributed along the circumferential direction on the outer surface of each impedance transformation segment. The microstructures occupy a portion of the length of the impedance transformation segment in the axial direction and maintain a first preset distance between the segments and the boundaries of adjacent segments. The geometric undulation of the microstructures is limited to at most a first predetermined proportion of the electromagnetic skin depth of the conductive material of the power divider at the upper limit frequency of the corresponding operating frequency band. The corners where the microstructures meet the power divider are rounded.

2. A power divider according to claim 1, characterized in that: The power divider includes at least a dense region of fine structures, which is a section in which the pitch density of the fine structures in the axial direction of the power divider is increased relative to the adjacent impedance transformation section. The inner wall of the cavity is provided with an impedance mode suppression structure, which is a plurality of annular grooves arrayed along the length direction on the inner wall of the cavity. The geometric dimensions of the annular grooves are limited to a first predetermined upper limit fraction of the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band, and the impedance mode suppression structure is arranged in an axial position opposite to the dense microstructure area on the power divider rod.

3. A power divider according to claim 2, characterized in that, The axial length of each impedance transformation segment is configured such that, within the dielectric space between the power divider and the cavity, the difference between the integrals of the electromagnetic energy density along the axial direction of the power divider, based on the field distribution at the upper limit frequency of the operating frequency band, of adjacent impedance transformation segments does not exceed a second predetermined upper limit ratio based on the larger of the two integrals.

4. A power divider according to claim 2, characterized in that, The end face of each output port forms an impedance trench along the circumferential direction; the geometric depth of the impedance trench is configured such that the equivalent electrical length of the impedance trench at the upper limit frequency of the operating frequency band approaches one-quarter of the guided wave wavelength of the TEM master mode. Furthermore, the impedance trench and the impedance suppression structure maintain a second preset distance in the axial direction.

5. A power divider according to claim 1, characterized in that, The impedance transformation segment is a multi-stage structure with a preset number of stages. The axial length of each stage of the impedance transformation segment is limited to a second predetermined upper limit fraction based on the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band. Furthermore, the ratio of the outer diameters of adjacent impedance transformation segments is limited to a deviation of no more than a third predetermined upper limit ratio relative to 1 and remains monotonically changing along the axial direction to achieve continuous impedance transition.

6. A power divider according to claim 2, characterized in that, The microstructure consists of circumferential shallow grooves periodically distributed along the circumference of the power divider; the pitch of each circumferential shallow groove is limited to a third predetermined upper limit fraction based on the wavelength of the TEM primary mode guided wave at the upper limit frequency of the operating frequency band; the geometric undulation of each circumferential shallow groove is limited to at most a second predetermined proportion of the electromagnetic skin depth of the conductive material of the power divider at the upper limit frequency of the operating frequency band; the groove edges of each circumferential shallow groove are rounded, and the transition between the microstructure and the power divider is rounded.

7. A power divider according to claim 4, characterized in that, The geometric dimensions of the annular groove include groove depth, groove width, and groove spacing; and are respectively limited to a fourth predetermined upper limit fraction, a fifth predetermined upper limit fraction, and a sixth predetermined upper limit fraction based on the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band. The width of the impedance trench is limited to no more than the corresponding seventh predetermined upper limit fraction relative to the wavelength of the TEM master mode guided wave at the upper limit frequency of the operating frequency band.

8. A power divider according to claim 1, characterized in that, Also includes: A slit cylindrical temperature-sensitive dielectric sheath is disposed on the outer surface of the power divider rod and located outside the compensation section determined by a temperature sensitivity criterion along the axial direction of the power divider rod. The temperature sensitivity criterion is based on the fact that the impedance and / or phase shift at the upper limit frequency of the operating frequency band changes with temperature to reach a first predetermined threshold. The temperature-sensitive dielectric sheath is made of a dielectric material with a non-zero dielectric constant temperature coefficient.

9. A power divider according to claim 8, characterized in that, Also includes: A circumferential resistive film is sleeved on the outside of the temperature-sensitive dielectric sheath and is only disposed within the low-field annular band determined by a node criterion along the axial direction of the power divider rod. The node criterion is based on the fact that the local electric field amplitude normalized to the preceding wave component at the upper limit frequency of the operating frequency band does not exceed a fourth predetermined upper limit ratio. Non-conductive segmented slits are provided in the circumferential direction of the circumferential resistive film, and a gradual end is provided at its axial edge.

10. A power divider according to claim 1, characterized in that: The power divider includes at least a dense microstructure region, and the inner wall of the cavity is provided with an impedance mode suppression structure. The mode suppression structure is an annular groove arrayed along the length direction, and its geometric dimensions are limited to a first predetermined upper limit fraction of the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band. The mode suppression structure is arranged in an axial position opposite to the dense microstructure region. The axial length of each impedance transformation segment is configured such that the difference between the integrals of the electromagnetic energy density along the axial direction of the power divider in the dielectric space between the power divider and the cavity, based on the field distribution at the upper limit frequency of the operating frequency band, does not exceed a second predetermined upper limit ratio based on the larger of the two integrals. Each of the output ports has an impedance trench formed on its end face along the circumferential direction, and its geometric depth is configured such that the equivalent electrical length at the upper limit frequency of the operating frequency band approaches one-quarter of the waveguide wavelength of the TEM main mode, and maintains a second preset distance from the impedance mode suppression structure in the axial direction. The impedance transformation segment is a multi-stage structure with a preset number of stages. The axial length of each stage is limited to a second predetermined upper limit fraction based on the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band. The ratio of the outer diameters of adjacent impedance transformation segments is limited to a deviation of no more than a third predetermined upper limit ratio relative to 1 and remains monotonically changing along the axial direction to achieve continuous impedance transition. The microstructure consists of circumferential shallow grooves periodically distributed along the circumference of the power divider; the pitch of each circumferential shallow groove is limited to a third predetermined upper limit fraction based on the wavelength of the TEM primary mode guided wave at the upper limit frequency of the operating frequency band; the geometric undulation of each circumferential shallow groove is limited to at most a second predetermined proportion of the electromagnetic skin depth of the conductive material of the power divider at the upper limit frequency of the operating frequency band; the groove edges of each circumferential shallow groove are rounded, and the transition between the microstructure and the power divider is also rounded. The geometric dimensions of the annular groove of the mode suppression structure include groove depth, groove width, and groove spacing, which are respectively limited to a fourth predetermined upper limit fraction, a fifth predetermined upper limit fraction, and a sixth predetermined upper limit fraction based on the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band; the width of the impedance trench is limited to a seventh predetermined upper limit fraction relative to the wavelength of the TEM main mode guided wave at the upper limit frequency of the operating frequency band. The power divider also includes: A slit cylindrical temperature-sensitive dielectric sheath is disposed on the outer surface of the power divider rod and located outside the compensation section determined by a temperature sensitivity criterion along the axial direction of the power divider rod. The temperature sensitivity criterion is based on the fact that the impedance and / or phase shift at the upper limit frequency of the operating frequency band changes with temperature to reach a first predetermined threshold. The temperature-sensitive dielectric sheath is made of a dielectric material with a non-zero dielectric constant temperature coefficient. A circumferential resistive film is sleeved on the outside of the temperature-sensitive dielectric thin sleeve, and is only set in the low-field annular band determined by the node criterion along the axial direction of the power divider rod. The node criterion is based on the fact that the local electric field amplitude normalized by the preceding wave component at the upper limit frequency of the operating frequency band does not exceed a fourth predetermined upper limit ratio. Non-conductive segmented slits are provided in the circumferential direction of the circumferential resistive film, and a gradual end is provided at its axial edge.

Images