A radio frequency mixer integrated with adaptive adjustment function

Abstract

The application discloses a radio frequency mixer integrated with self-adaptive adjusting function and belongs to the technical field of electronic circuits, comprising: the mixer comprises a waveguide main body, an integrated mixer assembly, an embedded parasitic compensation network and a self-adaptive control unit. The inner wall of the waveguide main body is provided with an annular groove, and a cylindrical shielding cover is embedded in the groove so that the inner surface of the shielding cover is coplanar with the inner wall of the waveguide, and the protrusion problem of the traditional shielding cover is eliminated. The mixing diode is accurately installed on the microstrip substrate at the bottom of the groove, and the distance between the two components is set to the odd multiple of the half wavelength of the local oscillator signal. The embedded parasitic compensation network adopts a three-dimensional spiral capacitor structure and is connected with the mixing diode through a micro ground via. The self-adaptive control unit monitors the performance index in real time through a sensor and dynamically adjusts the position of the movable short-circuit end and the diode bias by using a gradient algorithm. The application solves the problem of phase mismatch in the high-frequency band and significantly improves the conversion efficiency, isolation and linearity.

Description

Technical Field

[0001] This invention relates to the technical field of demodulating or transforming modulation from one carrier frequency to another, and more particularly to a radio frequency mixer with integrated adaptive adjustment function. Background Technology

[0002] As a core component in wireless communication, radar detection, and high-frequency signal processing systems, the radio frequency (RF) mixer plays a crucial role in nonlinearly mixing RF signals with local oscillator (LO) signals to achieve frequency conversion. With the evolution of communication protocols towards higher frequencies and larger bandwidths, more stringent requirements are placed on mixer performance, including conversion loss, isolation, and linearity. To improve performance, the industry has developed various technical solutions: one is based on a waveguide structure, which creates an electromagnetic standing wave field within it. The spacing between a pair of mixer transistors (such as diodes) is precisely set to an odd multiple of half the wavelength of the LO signal, allowing the LO signal to be fed in reverse phase. Adjusting the short-circuit terminal aligns the standing wave antinodes with the transistor input, achieving efficient energy coupling and difference frequency output. Another approach focuses on device optimization, coaxially mounting the mixer diodes within a conductive shield, combined with symmetrical cross-connections and compensation capacitors, effectively suppressing parasitic reactance, improving isolation, and reducing conversion loss. These solutions demonstrate significant advantages in independent applications, providing crucial support for improving the performance of high-frequency systems.

[0003] Despite the progress made in optimizing electromagnetic field distribution and improving device structure, these technologies present significant challenges when integrated to achieve synergistic effects. Existing integration schemes, when combining waveguide standing wave structures with shielding tube assemblies, suffer from a fixed radial displacement (measured up to 0.8 mm) of the core operating point of the mixer diode relative to the waveguide inner wall due to the inherent physical thickness of the shielding tube (typically not less than 1.2 mm). This displacement far exceeds the position tolerance (typically ±0.1 mm) in high-frequency applications (e.g., above 10 GHz), disrupting the phase matching conditions required for the waveguide standing wave field. The consequences include phase mismatch in the local oscillator signal feed, impaired symmetry of the balanced structure, significantly deteriorating the cancellation effect of local oscillator noise and spurious harmonics, ultimately leading to increased output signal distortion (measured up to 3.2 dB) and increased conversion loss. This deficiency stems from the inherent contradiction between the requirements for precise macroscopic electromagnetic field distribution and the realization of microscopic component physical structures, limiting the application of integrated schemes in high-performance scenarios. Summary of the Invention

[0004] This invention overcomes the shortcomings of the prior art and provides a radio frequency mixer with integrated adaptive adjustment function.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is: a radio frequency mixer with integrated adaptive adjustment function, comprising:

[0006] The waveguide body defines an electromagnetic propagation channel inside and is provided with a first port for feeding in a first frequency signal, a second port for feeding in a second frequency signal, a third port for deriving a difference frequency signal, and a movable short-circuit terminal.

[0007] At least two integrated mixer components are symmetrically arranged on the inner wall of the waveguide body with a preset spacing; wherein the preset spacing is set to be an odd multiple of half the wavelength of the first frequency signal inside the waveguide body.

[0008] An embedded parasitic compensation network is conformally integrated with the integrated mixer assembly within the wall structure of the waveguide body;

[0009] An adaptive standing wave and bias control unit establishes a control connection with the movable short-circuit terminal of the waveguide body and the integrated mixer assembly.

[0010] In a preferred embodiment of the present invention, each of the integrated mixer components includes:

[0011] An annular groove formed on the inner wall of the waveguide body;

[0012] A cylindrical shielding cover is embedded and fixed in the annular groove, wherein the inner surface of the cylindrical shielding cover is coplanar with the original inner wall surface of the waveguide body;

[0013] A multilayer microstrip substrate disposed at the bottom of the annular groove, wherein the top metal pattern of the multilayer microstrip substrate includes mixer diode mounting pads, intermediate frequency output connection lines, and parasitic compensation network microstrip lines; and

[0014] A mixer diode mounted on the multilayer microstrip substrate, wherein the active junction region of the mixer diode is positioned perpendicular to the inner wall of the waveguide.

[0015] The center-to-center distance between the two annular grooves is an odd multiple of half the wavelength of the first frequency signal within the waveguide.

[0016] In a preferred embodiment of the present invention, the inner surface of the waveguide body is chemically polished to reduce the surface roughness to Ra0.4μm, and a nickel transition layer and a pure silver conductive layer are deposited sequentially thereon; wherein the thickness of the nickel transition layer is 1-2μm and the thickness of the pure silver conductive layer is 4-5μm.

[0017] In a preferred embodiment of the present invention, the cylindrical shield is made of Kovar alloy and has a gold layer plated on its inner and outer surfaces; the mixer diode is a gallium arsenide Schottky beam-leaded diode.

[0018] In a preferred embodiment of the present invention, the embedded parasitic compensation network is disposed on the outer wall of the waveguide body and spatially aligned with the annular groove of the integrated mixer assembly. The embedded parasitic compensation network includes:

[0019] Three-dimensional spiral capacitor structure;

[0020] And a micro-grounding via penetrating the main wall of the waveguide;

[0021] In this configuration, one electrode of the three-dimensional spiral capacitor structure is connected to a port of the mixer diode within the integrated mixer assembly through the micro grounding via, while the other electrode is directly connected to the outer wall of the waveguide body, which serves as the system's common ground, to achieve a large-area, low-impedance connection.

[0022] In a preferred embodiment of the present invention, the three-dimensional spiral capacitor structure is formed by multilayer thin film deposition and photolithography, and is composed of multiple alternating stacked metal conductor layers and high dielectric constant dielectric layers; the micro grounding via is formed by laser drilling and filled with conductive silver paste, and its physical length is limited to the wall thickness of the waveguide body.

[0023] In a preferred embodiment of the present invention, the micro grounding via is formed by picosecond laser drilling process, with a hole diameter of 0.05-0.1 mm, and is filled with highly conductive silver paste; the highly conductive silver paste is composed of 85% silver powder, 10% epoxy resin and 5% reactive diluent, and the silver powder particle size distribution is 1-5 μm.

[0024] In a preferred embodiment of the present invention, the adaptive standing wave and bias control unit includes:

[0025] An actuator subunit is used to adjust the position of the movable short-circuit terminal and the DC bias voltage of the mixer diode;

[0026] The sensor subunit is used to monitor the key performance indicators of the RF mixer in real time; and

[0027] The processing and control unit receives the monitoring data from the sensor subunit and outputs control signals to the actuator subunit according to a preset algorithm.

[0028] In a preferred embodiment of the present invention, the actuator subunit includes:

[0029] A piezoelectric ceramic stack micro-displacement device mechanically coupled to the movable short-circuit end of the waveguide body, the piezoelectric ceramic stack micro-displacement device generating a linear displacement with nanometer-level resolution under the drive of a control signal, so as to fine-tune the position of the movable short-circuit end; and

[0030] A high-precision digital-to-analog converter is used to provide a precisely adjustable DC bias voltage to the mixer diode according to a control signal.

[0031] In a preferred embodiment of the present invention, the sensor subunit includes:

[0032] A dual directional coupler is placed near the first port to monitor the forward power and reflected power of the first frequency signal input to the mixer;

[0033] A power detector is installed at the third port for real-time measurement of the output power of the difference frequency signal; and

[0034] A signal analysis module for performing spectral analysis on the difference frequency signal and extracting the power values ​​of its third-order intermodulation distortion components.

[0035] This invention addresses the shortcomings of the prior art and has the following beneficial effects:

[0036] (1) This invention achieves dynamic tuning of the internal electromagnetic field distribution by adopting a waveguide body and a movable short-circuit terminal structure design, combined with an adaptive standing wave and bias control unit. By adjusting the position of the short-circuit terminal in real time, the antinodes of the standing wave are precisely moved, ensuring that the mixer diode is always at the optimal energy coupling point, thereby improving signal conversion efficiency and consistency. Compared with traditional waveguide mixers with fixed boundary conditions, which cannot adapt to performance degradation caused by frequency drift or environmental changes, this invention further enhances the frequency adaptability and environmental robustness of the mixer in high-frequency communication systems.

[0037] (2) This invention eliminates the radial displacement problem caused by the physical thickness of traditional shielding structures by integrating the annular groove of the integrated mixer assembly with the cylindrical shielding cover in a coplanar manner; it ensures that the active junction region of the mixer diode is precisely aligned with the inner wall of the waveguide, maintains the integrity and uniformity of the electromagnetic field distribution, and improves the anti-phase excitation effect of the local oscillator signal and the port isolation. Compared with the standing wave field distortion and phase mismatch caused by existing surface-mount shielding covers, this invention further reduces the risk of signal distortion and provides a stable balanced mixing foundation for high-frequency applications such as the Ka band.

[0038] (3) This invention integrates the compensation element directly into the waveguide wall through the three-dimensional spiral capacitor and miniature grounding via structure of the embedded parasitic compensation network, which greatly shortens the grounding loop path; by using the waveguide itself as a low-impedance ground wire, the self-resonance phenomenon caused by parasitic inductance is effectively suppressed, the operating bandwidth is expanded and the conversion loss characteristics are optimized. Compared with the traditional scheme of using surface-mount capacitors and long lead grounding, its high-frequency performance is severely limited. This invention further ensures the pure capacitive compensation and linearity improvement of the mixer in a wide frequency range.

[0039] (4) This invention constructs a closed-loop real-time optimization system through a collaborative mechanism between a high-resolution micro-displacer in the adaptive control unit and multiple sensors. Based on the gradient climbing algorithm, the phase of the standing wave field and the diode bias are dynamically adjusted to achieve synchronous optimization of conversion gain and linearity. Compared with traditional manual adjustment or fixed bias schemes that cannot respond to dynamic operating conditions, this invention further enhances the performance stability and overall efficiency of the mixer under complex scenarios such as temperature fluctuations and power changes. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a simplified circuit diagram of a preferred embodiment of the present invention;

[0042] Figure 2 This is a timing diagram of the signal processing flow of a preferred embodiment of the present invention;

[0043] Figure 3 This is a timing diagram of the adaptive control mechanism of a preferred embodiment of the present invention;

[0044] In the figure: 1. Waveguide body; 2. First port; 3. Second port; 4. Integrated mixer assembly; 5. Movable short-circuit terminal; 6. Embedded parasitic compensation network; 7. Adaptive standing wave and bias control unit. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein. Therefore, the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0047] Application Overview:

[0048] This application addresses waveguide RF mixers with integrated shielded tube structures. Existing technologies are limited by their inability to overcome the uncontrollable interference caused by the physical dimensions of the shielding components on the electromagnetic field distribution within the waveguide. When a shielding structure is integrated inside a waveguide, its physical presence inevitably alters the waveguide's boundary conditions, thereby disturbing the precise standing wave field distribution. In principle, the positions of the antinodes and nodes of the standing wave field within the waveguide, crucial for efficient mixing, are determined by both the electromagnetic wave propagation characteristics and the boundary conditions. The introduced shielding component, with its metal walls forming a new, local electromagnetic boundary, scatters the incident wave, distorting the original ideal standing wave distribution. Consequently, the mixer diode, pre-positioned at the antinode for maximum energy coupling, is actually located in a position where the electric field strength is weakened and the phase is mismatched.

[0049] This phase mismatch introduced by the fixed structure fundamentally disrupts the symmetry principle upon which balanced mixing relies. The local oscillator signal cannot excite a pair of mixer diodes in a precisely out-of-phase manner, causing the local oscillator leakage suppression mechanism to fail and reducing the in-phase superposition efficiency of the intermediate frequency signals generated by mixing. Therefore, the decrease in energy coupling efficiency and signal distortion are not accidental, but an inevitable result of the electromagnetic field distribution being disrupted by the rigid physical structure.

[0050] The limitation of existing technological approaches lies in attempting to find a suboptimal solution through fine-tuning in a disturbed, fixed electromagnetic environment. However, once the operating frequency of the waveguide changes, the wavelength and standing wave distribution within it also change, and the original optimal installation location will quickly become mismatched at the new frequency.

[0051] Based on this fundamental understanding, the breakthrough of this application lies in no longer presupposing a static structure applicable to all operating conditions, but instead constructing a system capable of dynamically responding and actively reconstructing the internal electromagnetic field distribution. By introducing movable boundary conditions, the waveguide body is transformed from a passive signal transmission channel into an actively tunable resonant structure.

[0052] Exemplary device:

[0053] A radio frequency mixer with integrated adaptive adjustment function, comprising:

[0054] The waveguide body defines an electromagnetic propagation channel inside and is provided with a first port for feeding in a first frequency signal, a second port for feeding in a second frequency signal, a third port for deriving a difference frequency signal, and a movable short-circuit terminal.

[0055] At least two integrated mixer components are symmetrically arranged on the inner wall of the waveguide body with a preset spacing; wherein the preset spacing is set to an odd multiple of half the wavelength of the first frequency signal inside the waveguide body.

[0056] An embedded parasitic compensation network is conformally integrated with the integrated mixer assembly within the wall structure of the waveguide body;

[0057] An adaptive standing wave and bias control unit establishes a control connection with the movable short-circuit terminal of the waveguide body and the integrated mixer assembly.

[0058] In the frequency conversion and energy coupling links of radio frequency signals, waveguide structures, as the basic carriers for electromagnetic field confinement, signal transmission, and the installation of core components, not only need to meet the basic requirements of low-loss propagation of high-frequency signals, but also need to provide physical adaptation and electromagnetic compatibility guarantees for the deep integration of multifunctional modules. However, existing technologies face a fundamental contradiction when attempting to integrate high-isolation shielding structures: the physical thickness of the shielding cover introduced to achieve electromagnetic shielding causes a fixed radial displacement of the active mixer devices inside the waveguide relative to the inner wall of the waveguide. This displacement far exceeds the position tolerance at high frequencies, directly disrupting the precise standing wave field distribution inside the waveguide, leading to phase mismatch and performance degradation.

[0059] The waveguide body is not only a physical channel for the directional propagation of high-frequency electromagnetic energy, but also a dynamically tunable multimode resonant structure, providing precise electromagnetic boundary conditions for the injection, synthesis, and extraction of signals of different frequencies. This invention, by designing the waveguide body as an active resonant device with movable boundary conditions, overcomes the shortcomings of traditional fixed structures that cannot adapt to changing operating conditions. This allows the mixer to dynamically adjust its internal electromagnetic field distribution according to real-time operating conditions, ensuring optimal signal coupling efficiency under any circumstances.

[0060] Unlike conventional waveguides in the prior art that only serve as signal channels or provide static standing wave fields, the waveguide main structure of this invention integrates multiple functional ports and a key movable terminal, which together constitute the physical basis for actively intervening in and reconstructing the internal field distribution.

[0061] In one specific embodiment, the waveguide body is a standard WR-28 rectangular waveguide structure suitable for Ka-band operation, with an internal channel cross-sectional dimension of 7.112mm × 3.556mm, the dimensions of which strictly comply with the physical characteristics requirements of electromagnetic wave propagation in this frequency band; the Ka-band is 26.5-40GHz;

[0062] The waveguide body is made of oxygen-free high-conductivity copper C10100 as the substrate and is manufactured in one piece with a dimensional tolerance of ±0.01mm.

[0063] To further optimize high-frequency signal transmission performance, surface treatment is performed inside the waveguide body. In the Ka band and higher frequencies, the skin effect causes the current to mainly flow on the conductor surface, and surface roughness becomes a key factor affecting transmission loss.

[0064] The surface treatment includes: using a chemical polishing process to reduce the surface roughness to Ra0.4μm, and then using an electroplating process to sequentially deposit a 1-2μm nickel transition layer and a 4-5μm pure silver conductive layer; the composite surface not only ensures excellent conductivity but also provides durable surface protection.

[0065] Chemical polishing eliminates microscopic irregularities and subsurface damage layers left by machining through controlled chemical corrosion. Specifically, the waveguide body is immersed in an acidic polishing solution at 45-55°C for 1.5-3 minutes. The solution utilizes its higher dissolution rate at microscopic protrusions in the metal to achieve micro-leveling and mirror-like finish. The acidic polishing solution is a phosphoric acid-based mixed solution with the following volume ratios: phosphoric acid 60%-70%, nitric acid 5%-8%, acetic acid 10%-15%, sulfuric acid 5%-10%, and the addition of 0.5%-1.5% corrosion inhibitor and 0.1%-0.5% nonionic surfactant.

[0066] After chemical polishing and thorough cleaning and activation, electrochemical deposition of multilayer metals is performed to construct a composite surface structure;

[0067] The nickel transition layer is deposited using the nickel sulfamate electroplating process, at a temperature of 47-53℃ and a cathode current density of 2-4 A / dm³. 2 The plating process is carried out under the following conditions: the pH value of the plating bath is between 3.8 and 4.2, and the deposition time is within 5-10 minutes, which can form a dense nickel transition layer with a thickness of 1-2 μm. As a key functional barrier layer, the nickel transition layer can effectively inhibit the interdiffusion between the copper atoms in the substrate and the subsequent silver layer, and provide an excellent adhesion basis for the pure silver conductive layer.

[0068] The deposition of a pure silver conductive layer was carried out using a potassium silver cyanide alkaline electroplating system at a relatively low temperature of 22-28℃, with the cathode current density controlled at 0.8-1.2 A / dm³. 2 Within a certain range; the deposition time is 15-25 minutes, ultimately obtaining a pure silver conductive layer with a thickness of 4-5 μm. The pure silver conductive layer, with its extremely low bulk resistivity, becomes the main pathway for high-frequency current. Its low-stress, dense structure directly interacts with the propagating electromagnetic field, thus providing crucial surface conductivity for achieving low transmission loss in the millimeter-wave band.

[0069] The waveguide body is equipped with three flange connection ports conforming to the UG-599 / U standard, each of which has a clearly defined signal interface function.

[0070] The first port is used to receive the local oscillator signal injection from an external signal source, i.e., the first frequency signal; the second port is used to output the intermediate frequency signal after mixing, i.e., the second frequency signal; and the third port is used to export the intermediate frequency signal, i.e., the difference frequency signal.

[0071] One end of the waveguide body is provided with a movable short-circuit terminal. The movable short-circuit terminal adopts a non-contact choke groove structure to avoid metal particles and unstable contact resistance that may be generated due to physical contact, and to ensure that it has a reflection coefficient close to 1 at high frequencies.

[0072] The waveguide body, through its integrated structure, surface treatment process, and movable boundary design, lays the physical foundation for achieving high-performance signal synthesis and adaptive modulation. Based on this, an integrated mixer assembly is introduced to achieve the positioning and integration of the mixer device and electromagnetic field distribution.

[0073] Integrated mixer components are used to fundamentally eliminate physical position deviations introduced by the installation of mixer devices. In high-frequency scenarios such as the Ka band, position deviations of mixer devices will directly cause them to deviate from the energy-rich region of the standing wave field, resulting in a significant decrease in energy coupling efficiency and phase mismatch. Traditional structures are difficult to compensate for this through later adjustments, so this deviation must be completely eliminated from the source of component design.

[0074] Each integrated mixer assembly includes: an annular groove formed on the inner wall of the waveguide body, a cylindrical shield embedded and fixed in the annular groove, a multilayer microstrip substrate disposed at the bottom of the annular groove, and a mixer diode mounted on the multilayer microstrip substrate.

[0075] Among them, the annular groove is a precision structure formed on the inner wall of the waveguide body by micro milling cutter or femtosecond laser etching process. It is symmetrically arranged along the length of the waveguide, and its center spacing and size parameters are strictly matched with the distribution law of the standing wave field of the local oscillator signal. The center spacing of the two annular grooves is set to be an odd multiple of half the wavelength of the local oscillator signal, i.e. the first frequency signal, in the waveguide.

[0076] Because the local oscillator signal will form a standing wave field when it propagates in the waveguide, the antinodes are the points where the electric field intensity is at its maximum. The interval between the two mixer components is an odd multiple of half the wavelength. Placing the two mixer components at this interval can ensure that the two mixer diodes are simultaneously at the antinodes of the standing wave field, thereby maximizing energy coupling. Furthermore, the symmetrical distribution can cancel out signal phase deviation and improve mixing consistency.

[0077] For example, for a local oscillator signal with a center frequency of 35 GHz, since the guided wave wavelength inside the WR-28 waveguide is approximately 11.8 mm, the center-to-center distance between the two grooves is set to 1.5 times the guided wave wavelength, i.e., 17.7 mm. The groove diameter is 1.9-2.0 mm, and the depth is 0.4-0.5 mm, with a depth tolerance of no more than ±0.005 mm. This provides a reference positioning for subsequent component installation. The setting of the diameter and depth takes into account the size matching requirements of the cylindrical shield and the multilayer microstrip substrate. It is necessary to ensure the stable installation of each component without occupying too much internal space of the waveguide to avoid affecting the electromagnetic field distribution.

[0078] A cylindrical shield, made of Kovar alloy, is embedded and fixed within each annular groove. The material composition includes 27-29% nickel, 15-17% cobalt, and 54-58% iron, possessing a thermal expansion coefficient similar to that of the waveguide body. This prevents structural deformation due to temperature changes. Kovar alloy was chosen because the waveguide body uses oxygen-free, high-conductivity copper, whose thermal expansion coefficient is close to that of Kovar alloy, reducing thermal stress deformation under high and low temperature environments and preventing gaps between the shield and the grooves, which could compromise the electromagnetic shielding effect and coplanarity. Both the inner and outer surfaces of the cylindrical shield are plated with a 5μm thick gold layer to ensure excellent conductivity and solderability. The gold layer not only reduces the bulk resistivity of the shield, decreasing high-frequency signal loss, but also improves surface solderability and oxidation resistance, ensuring long-term operational stability. The outer diameter of the cylindrical shield precisely matches the diameter of the annular groove, with a wall thickness of 0.13-0.15mm, and its height perfectly matches the depth of the annular groove, ensuring no protrusions or gaps after assembly.

[0079] The cylindrical shield is permanently fixed in the groove by laser welding or conductive silver paste. The installation accuracy requires that its inner surface be coplanar with the original inner wall surface of the waveguide body, and the flatness deviation between the two is within ±0.003 mm. Among them, laser welding has higher connection strength, while conductive silver paste is more suitable for mass assembly.

[0080] Embedded design fundamentally eliminates the protruding structure formed inside the waveguide by the physical thickness of traditional surface-mount shields, thereby ensuring the integrity and uniformity of the electromagnetic field inside the waveguide and laying the physical foundation for achieving an ideal standing wave field distribution.

[0081] At the bottom of the annular groove, a multilayer microstrip substrate is fixed. The multilayer microstrip substrate is made of Rogers RO4350B high-frequency laminate material with a dielectric constant of 3.48 and a thickness of 0.254 mm.

[0082] The top metal pattern of the multilayer microstrip substrate includes: mixer diode mounting pads, intermediate frequency output connection lines, and parasitic compensation network microstrip lines; the positions of each pattern are optimized according to the standing wave field distribution and signal transmission path to ensure impedance matching and signal integrity.

[0083] For example, the mixer diode is a gallium arsenide (GaAs) Schottky beam-lead device, specifically the MA4E2038 series from MACOM. It has a cutoff frequency ≥140GHz, zero-bias junction capacitance ≤0.02pF, and series resistance ≤3Ω, exhibiting high-efficiency mixing and conversion characteristics. GaAs material has high electron mobility, much higher than silicon material, which can meet the high-frequency response requirements of the Ka band. The high cutoff frequency, low junction capacitance, and low series resistance of the MA4E2038 series can reduce the signal loss of the device itself and improve the mixing and conversion efficiency.

[0084] The bonding pads on the multilayer microstrip substrate are installed using a thermo-ultrasonic bonding process, with the bonding temperature controlled at 180-200℃ and the ultrasonic power at 0.8-1.2W. During the installation process, a visual alignment system with a magnification of ≥200x is used to assist in positioning, ensuring that the positional deviation of the core active junction region in the direction perpendicular to the inner wall of the waveguide is within a tolerance of ±0.05mm.

[0085] This ensures that the active junction region of the diode can be precisely placed at the antinode of the standing wave field formed by the local oscillator signal and the movable short-circuit terminal, i.e. the point of maximum electric field strength, thereby achieving maximum energy coupling and the most efficient mixing conversion.

[0086] The cylindrical shield is made of highly conductive metal material, providing a local and independent electromagnetic shielding environment for the mixer diode to suppress stray reactance and crosstalk between ports. The independent shielding environment is set up because intermediate frequency signals and stray harmonics are generated during the mixing process. Signals from different ports are prone to crosstalk through electromagnetic field coupling. The shield can isolate external electromagnetic interference and suppress stray reactance of the mixer diode to avoid affecting the uniformity of the standing wave field in the waveguide.

[0087] The integrated mixer assembly, through the combination of multiple structures—including the reference positioning of the annular groove, the coplanar design of the cylindrical shield, the precise support of the multi-layer microstrip substrate, and the precision mounting of the mixer diodes—constructs a multi-dimensional, high-precision integrated system. This multi-structure combination design is adopted because a single structure cannot simultaneously solve the three core problems of positional deviation, electromagnetic shielding, and energy coupling. The annular groove provides the reference, the shield ensures the integrity of the electromagnetic field, the substrate achieves signal transmission and impedance matching, and the diodes complete energy conversion. The synergistic effect of these structures enables high-performance mixing. This not only fundamentally solves the positional deviation problem of traditional mixer components but also suppresses stray reactance and port crosstalk through electromagnetic shielding. It provides core structural support for the RF mixer to achieve low insertion loss, high isolation, and high conversion efficiency, and is the key execution unit for achieving precise standing wave field matching in this invention.

[0088] Building upon the precise standing wave field matching achieved by integrated mixer components, embedded parasitic compensation networks serve as a key impedance control module in high-frequency links. This minimizes parasitic inductance in the grounding path, preventing self-resonance at high frequencies from degrading mixing performance. In traditional balanced mixers, the grounding path of cross-coupled capacitors is typically connected using leads or surface microstrip lines. In the Ka band and higher frequencies, this long path introduces significant parasitic inductance, causing the compensation capacitors to exhibit inductive characteristics, triggering self-resonance, worsening the mixer's conversion losses, narrowing bandwidth, and severely limiting the upper limit of high-frequency performance.

[0089] The embedded parasitic compensation network integrates the compensation capacitor and grounding via directly into the waveguide body wall. It utilizes the waveguide itself as a low-impedance common ground line and achieves pure capacitive compensation over a wide range by shortening the grounding loop length and optimizing the capacitor structure.

[0090] The embedded parasitic compensation network and the integrated mixer assembly are spatially precisely aligned and integrated within the wall structure of the waveguide body.

[0091] The embedded parasitic compensation network includes: a three-dimensional spiral capacitor structure and a miniature grounding via;

[0092] Among them, a three-dimensional spiral capacitor structure was constructed on the outer wall of the waveguide body, corresponding to the position of each annular groove, using thin film manufacturing process;

[0093] Specifically, the three-dimensional helical capacitor structure is formed by alternating magnetron sputtering of metal thin films and plasma-enhanced chemical vapor deposition of dielectric thin films;

[0094] By repeatedly depositing and photolithographic etching processes, a multilayer stacked structure with 8 metal conductor layers and 7 dielectric isolation layers was constructed and patterned into a planar spiral shape to form a spiral capacitor with 16 layers, thereby achieving a precise capacitance value of about 0.5 picofarads in a very small footprint.

[0095] One electrode of the three-dimensional spiral capacitor structure is directly connected to a port of the mixer diode in the integrated mixer assembly through a miniature grounding via. This direct connection design shortens the signal path, reduces additional parasitic parameters, and ensures that the parasitic reactance of the compensation capacitor and the mixer diode cancels out in real time. The other electrode is directly connected to the outer wall of the waveguide body, which serves as the common ground, to achieve a large-area low-impedance connection. The waveguide body, as the common ground, has a resistance close to zero. The large-area connection further reduces the grounding impedance and avoids grounding potential drift.

[0096] For example, the fabrication of a three-dimensional spiral capacitor structure begins in the substrate preparation stage.

[0097] In a designated area on the outer wall of the waveguide body that has undergone surface treatment, surface contaminants are first removed by plasma cleaning, followed by the deposition of a 200nm thick silicon dioxide insulating layer. The insulating layer serves to prevent unnecessary electrical connections between the capacitor structure and the waveguide body, while also providing a flat substrate for subsequent thin film deposition. Silicon dioxide was chosen because of its excellent insulating properties and adhesion to metal thin films.

[0098] A magnetron sputtering process was employed, first depositing a 20-25 nm thick titanium film as an adhesion layer on a silicon dioxide insulating layer, followed by a 500-550 nm thick gold film as a conductor layer. The titanium adhesion layer was chosen because of its excellent adhesion to both silicon dioxide and gold, preventing gold layer detachment. The gold conductor layer was selected due to its superior conductivity and oxidation resistance. During deposition, the substrate temperature was maintained at 145-150 °C, and the sputtering pressure at 0.5 Pa to ensure the density and uniformity of the film.

[0099] After the first metal layer is deposited, the dielectric layer is prepared. A silicon nitride dielectric layer with a thickness of 180-200 nm is deposited on the metal layer using plasma-enhanced chemical vapor deposition. The reaction gas is a mixture of silane and ammonia, the deposition temperature is controlled at 250-300℃, and the radio frequency power is 100W. Silicon nitride is chosen because it has a high dielectric constant and good insulation strength, which can provide a larger capacitance per unit area at the same thickness.

[0100] The alternating metal-dielectric deposition cycle was then repeated. Each cycle consisted of two steps: metal layer deposition and dielectric layer deposition, with a total of 8 metal layer depositions and 7 dielectric layer depositions. Before each metal layer deposition, a specific spiral pattern was defined using photolithography and etching processes, employing positive photoresist and a specific developer. After exposure, a gold-specific etchant was used for patterning. This repeated process ultimately formed a three-dimensional spiral structure with 16 functional layers.

[0101] After all layers are stacked, final patterning and passivation are performed. Electrode contact windows are defined using photolithography, and reactive ion etching is used to open the top dielectric layer. The exposed metal electrode area is then thickened and metallized. Finally, a 250-300 nm thick silicon nitride passivation layer is deposited on the entire structure surface to prevent external environmental erosion and damage to the capacitor structure.

[0102] The miniature grounding via is formed on the wall of the waveguide body using picosecond laser drilling technology, with a hole diameter of 0.05-0.1mm; the laser wavelength is 355-370nm, the pulse width is 10ps, the repetition frequency is 80-100kHz, and the single pulse energy is 50μJ; nitrogen-assisted purging is used during the drilling process to remove machining debris in time and prevent oxidation.

[0103] After drilling, the inner wall of the micro grounding via is metallized and activated with oxygen plasma at a power of 300W for 5 minutes to increase the surface energy of the via. It is then completely filled with highly conductive silver paste and cured at 150°C for 30 minutes.

[0104] The highly conductive silver paste consists of 85% silver powder, 10% epoxy resin and 5% reactive diluent. The silver powder particle size distribution is controlled at 1-5μm to ensure good flowability and dense filling.

[0105] The filling process is carried out in a vacuum environment. First, the waveguide body is preheated to 75-80°C. Then, a precision dispensing machine is used to slowly inject silver paste from the bottom of the hole upwards at a pressure of 0.2 MPa to avoid the formation of air bubbles.

[0106] After filling, a centrifugal leveling process is used, rotating the waveguide body at a speed of 2000 r / min to ensure that the silver paste surface is flush with the outer wall of the waveguide.

[0107] Because the physical length of the miniature grounding via is strictly limited to the thickness of the waveguide wall, the length of the entire grounding loop is greatly shortened.

[0108] According to the microstrip line inductance estimation formula, the parasitic inductance of this structure can be effectively suppressed to below 0.1 nanohenries. This low parasitic inductance ensures that the three-dimensional spiral capacitor structure exhibits pure capacitive characteristics in a frequency range up to 100 GHz, with its self-resonant frequency much higher than the mixer's operating frequency band. This guarantees that the mixer can maintain excellent conversion loss and port isolation performance even at high frequencies.

[0109] The embedded parasitic compensation network constructs a high-precision, low-loss parasitic compensation system through a high-capacitance density design of three-dimensional spiral capacitors, a short-path, low-inductance design of miniature grounding vias, and a conformal integration design with integrated mixer components. It not only fundamentally solves the parasitic inductance problem of traditional compensation structures, but also ensures the stability of the mixer's conversion loss and port isolation across the entire operating frequency band through its broadband pure capacitive characteristics. This provides crucial impedance control support for RF mixers to achieve wideband, high-performance frequency conversion.

[0110] Based on the integrated mixer components to achieve precise static standing wave field matching and the embedded parasitic compensation network to suppress parasitic parameters, the adaptive standing wave and bias control unit serves as the core of the entire mixer's control. Through closed-loop real-time optimization, it solves the performance degradation problems caused by dynamic factors such as temperature changes, input signal power fluctuations, and local oscillator frequency drift in the actual working environment.

[0111] The adaptive standing wave and bias control unit integrates high-precision actuators, multi-dimensional sensors, and high-speed processing units. With the comprehensive performance evaluation function as the optimization target, it dynamically adjusts the phase of the standing wave field and the operating point of the device through a gradient climbing algorithm to build an adaptive control mechanism, ensuring that the mixer always maintains the best comprehensive performance in all operating scenarios.

[0112] The adaptive standing wave ratio and bias control unit of this invention achieves multi-dimensional collaborative optimization, while taking into account the two core indicators of conversion gain and linearity. Through a unified performance evaluation function and synchronous control strategy, it avoids the performance degradation caused by optimizing a single indicator and achieves the global optimality of comprehensive performance.

[0113] The adaptive standing wave and bias control unit includes: an actuator subunit for physically adjusting the mixer's operating state, a sensor subunit for real-time monitoring of key performance indicators, and a processing and control unit as the intelligent core; the three are connected by signal and control links to form a closed-loop system, realizing full-process automation from performance perception to state regulation.

[0114] In one specific embodiment, the adaptive standing wave and bias control unit is adapted to the Ka-band WR-28 waveguide mixer, and its key structure, materials, and process parameters are as follows:

[0115] The actuator subunit includes: a piezoelectric ceramic stacked micro-displacement device mechanically coupled to the movable short-circuit end of the waveguide body, and a high-precision digital-to-analog converter for adjusting the DC bias voltage of the mixer diode.

[0116] The piezoelectric ceramic stacked micro-displacement device can generate linear displacement with nanometer-level resolution under the drive of an external control voltage, thereby fine-tuning the position of the movable short-circuit end and accurately moving the position of the local oscillator signal standing wave antinode inside the waveguide body.

[0117] For example, a dual-actuator design is adopted because it is necessary to simultaneously control the phase of the standing wave field and the operating point of the device. Only through their synergistic effect can optimal overall performance be achieved. Among them, the piezoelectric ceramic stacked micro-displacement device uses the P-841.10 model, which is rigidly connected to the movable short-circuit end of the waveguide body through a precision ceramic push rod. This model of micro-displacement device can generate a linear stroke of 15μm under a DC drive voltage of 0-100V, with a resolution of up to 0.3nm, which can meet the nanometer-level fine adjustment requirements of the standing wave antinode. Traditional electromagnetic displacement devices cannot achieve such a high resolution. The high-precision digital-to-analog converter uses the AD5761R to provide a precisely adjustable DC bias voltage for the two mixer diodes. The 16-bit resolution can achieve microvolt-level voltage adjustment, ensuring the accurate optimization of the nonlinear operating point of the mixer diodes and avoiding linearity deterioration caused by bias voltage deviation.

[0118] The sensor subunit includes a WR-28 dual directional coupler, a broadband power detector, and a signal analysis module. Multi-sensor collaborative monitoring is used to comprehensively acquire the mixer's operating status, as a single sensor cannot simultaneously cover input matching, output gain, and linearity indicators. The dual directional coupler is integrated near the first port, with a coupling degree of -20dB to -18dB. This coupling degree ensures that the coupled signal strength meets the detection requirements without causing significant loss to the main signal transmission. The forward and reverse signals coupled out are sent to the logarithmic power detector to monitor the local oscillator signal input power and port reflection loss, and to evaluate the input matching status in real time.

[0119] A broadband power detector is located at the third port output for real-time measurement of the intermediate frequency signal output power. The signal analysis module integrates a high-Q bandpass filter, a low-noise amplifier, and a high-sensitivity power detector. The center frequency of the bandpass filter is set at the frequency of the third-order intermodulation distortion product. The high-Q design can accurately separate the IMD3 component from the useful intermediate frequency signal, while the low-noise amplifier improves the detection sensitivity of weak distortion signals, ensuring the accurate quantification of linearity indicators.

[0120] The processing and control unit uses a 32-bit ARM Cortex-M7 microcontroller as its hardware core, specifically the STM32H7 series. This series of microcontrollers is characterized by its integrated hardware floating-point arithmetic unit and DSP instruction set, with a clock frequency of up to 200MHz, enabling high-speed processing of real-time data streams from multiple sensors. Traditional 8-bit or 16-bit microcontrollers cannot meet the speed requirements.

[0121] The microcontroller internally embeds a maximum performance tracking algorithm based on gradient climbing and defines a comprehensive performance evaluation function. This is the ratio of intermediate frequency output power to third-order intermodulation distortion power, where... This is the intermediate frequency output power. For order intermodulation distortion power, A higher value indicates better overall performance of the mixer; by integrating conversion gain and linearity, the problems caused by optimizing a single index are avoided.

[0122] The specific execution flow of the control algorithm is as follows: The microcontroller periodically applies a small positive perturbation to the driving voltage of the piezoelectric ceramic stack micro-displacement device, measures the performance change, then applies a negative perturbation, measures the performance change under the negative perturbation, determines the gradient direction based on the performance change after the two perturbations, and updates the algorithm according to the update rules. Adjust the drive voltage, where, This is the driving voltage for the piezoelectric ceramic at the next moment. The current driving voltage, To preset the step size factor, This represents the gradient value of the performance evaluation function.

[0123] The update rule locks the position of the movable short-circuit terminal at the optimal position corresponding to the maximum value of J. The gradient climbing algorithm can quickly converge to the local optimum, with a fast response speed, which is suitable for real-time control scenarios. At the same time, the same algorithm is used to synchronously optimize the DC bias voltage of the mixer diode to achieve fine tuning of the operating point. The synchronous control design is because the phase of the standing wave field and the operating point of the device are mutually influential, and adjusting a single parameter alone cannot achieve the global optimum.

[0124] The adaptive standing wave and bias control unit constructs a closed-loop adaptive control system through nanometer-level control of the actuator subunit, multi-dimensional perception and processing of the sensor subunit, and intelligent decision-making of the control unit. It not only fundamentally solves the defect of traditional mixers that cannot adapt to dynamic working environments, but also achieves synchronous optimization of conversion gain and linearity through comprehensive performance evaluation functions and collaborative control strategies.

[0125] like Figure 2 , Figure 3 As shown, the working process of the mixer includes:

[0126] After the system is initially powered on, the processing and control unit first initiates a self-calibration program. This unit sends control commands to the piezoelectric ceramic stacked micro-displacement device in the actuator subunit, driving it to complete a slow scan across its entire stroke range. Simultaneously, the dual-directional coupler in the sensor subunit monitors the reflected power data at the first port in real time and feeds it back to the processing and control unit. The processing and control unit analyzes the feedback data, locates the position of the micro-displacement device corresponding to the minimum reflected power, and sets this position as the approximate initial position of the movable short-circuit end, laying the foundation for subsequent precise operation.

[0127] After entering normal operation, the local oscillator signal is fed into the waveguide body through the first port, and the radio frequency (RF) signal is fed into the waveguide body through the second port. Because the integrated mixer components are symmetrically arranged at a preset spacing, this spacing causes the local oscillator signal to excite the two mixer diodes in an inverted manner, while the RF signal arrives at the mixer diodes in an in-phase manner. The two signals are frequency-mixed at the mixer diodes, and the resulting intermediate frequency (IF) signal is superimposed in phase at the third port and then output. The local oscillator signal and its even harmonics cancel each other out due to their inverted characteristics, effectively suppressing local oscillator leakage.

[0128] During normal mixing, the adaptive standing wave and bias control system operates continuously. The processing and control unit executes the maximum performance tracking algorithm at a fixed control cycle: first, a small positive disturbance is applied to the driving voltage of the piezoelectric ceramic stack micro-displacement device, and then the change in the performance evaluation function is acquired through the sensor subunit; then, an equal amount of negative disturbance is applied, and the corresponding change is acquired again. By analyzing the amplitude and sign of the two changes, the processing and control unit quickly calculates the gradient direction of the performance evaluation function with respect to the driving voltage, and then adjusts the driving voltage according to the preset gradient update rule to precisely lock the movable short-circuit terminal at the optimal position corresponding to the peak value of the performance evaluation function.

[0129] Meanwhile, the processing and control units employ the same gradient ramp logic to synchronously perturb and optimize the DC bias voltage of the mixer diodes in the integrated mixer assembly.

[0130] By using this dual-coordinated real-time control of the standing wave field phase based on the movable short-circuit terminal position adjustment and the diode operating point based on the bias voltage adjustment, a closed-loop feedback mechanism is formed. This mechanism can dynamically compensate for performance deviations caused by factors such as changes in ambient temperature and fine-tuning of the operating frequency, ensuring that the mixer always maintains its optimal operating state.

[0131] Example 1:

[0132] This embodiment provides a specific implementation method and performance verification of a Ka-band mixer constructed according to the technical solution of the present invention.

[0133] The specific structure of the mixer is as follows:

[0134] The waveguide body is made of OFHC copper and is machined into WR-28 specification by five-axis CNC. The inner wall is silver-plated with a thickness of 5μm.

[0135] The center-to-center distance between the two integrated mixer components is set to 17.70 mm, with the annular groove having a diameter of 2.0 mm and a depth of 0.50 mm.

[0136] The cylindrical shield is made of Kovar alloy, with a height of 0.50 mm and a wall thickness of 0.15 mm, and is fixed in the groove with H20E conductive silver paste.

[0137] The multilayer microstrip substrate is made of 0.254mm thick Rogers RO4350B material, and the mixer diode mounted on it is a MA4E2038-1141T beam-leaded diode.

[0138] In the embedded parasitic compensation network, the three-dimensional spiral capacitor structure is constructed by alternating deposition of 8 layers of gold conductors and 7 layers of silicon nitride dielectric, with each gold conductor layer being 500 nm thick and the dielectric layer being 200 nm thick, forming a capacitance value of 0.5 pF; the micro grounding via has a diameter of 100 μm and a length of 0.3 mm, and is filled with DuPont CB100 conductive silver paste.

[0139] The core of the adaptive standing wave and bias control system is the STM32H743 microcontroller, and the piezoelectric ceramic stacked micro-displacement device is the PI P-841.10 model.

[0140] During testing, the local oscillator signal was set to a frequency of 35.0 GHz and a power of +13 dBm. The radio frequency (RF) signals used were two-tone signals with frequencies of 35.10 GHz and 35.101 GHz, with a single-tone power of -10 dBm. The intermediate frequency (IF) output signals were 100 MHz and 101 MHz, corresponding to third-order intermodulation products at 99 MHz and 102 MHz. Testing was first conducted at a constant temperature of 25°C, and initial performance data were recorded. Subsequently, the ambient temperature was raised to 65°C, and performance parameters were collected again after the system stabilized to simulate device performance under harsh operating conditions.

[0141] Example 2:

[0142] The similarities with Example 1 will not be repeated here. The difference is that the depth of the annular groove of the integrated mixer assembly is 0.47 mm, the height of the cylindrical shield is 0.50 mm, and the inner surface of the shield protrudes 0.03 mm from the original inner wall of the waveguide body after assembly.

[0143] Example 3:

[0144] The similarities with Example 1 will not be repeated here. The difference is that the depth of the annular groove of the integrated mixer assembly is 0.45 mm, the height of the cylindrical shield is 0.50 mm, and the inner surface of the shield protrudes 0.05 mm from the original inner wall of the waveguide body after assembly.

[0145] Example 4:

[0146] The similarities with Example 1 will not be repeated here. The difference is that the three-dimensional spiral capacitor structure of the embedded parasitic compensation network adopts 6 metal conductor layers + 5 dielectric isolation layers. The thickness of each metal conductor layer is 500nm and the thickness of each dielectric layer is 200nm. The final capacitance value is 0.4pF. The structural parameters of the micro grounding via are the same as those of Example 1.

[0147] Example 5:

[0148] The similarities with Example 1 will not be repeated here. The difference is that the three-dimensional spiral capacitor structure of the embedded parasitic compensation network adopts 10 metal conductor layers + 9 dielectric isolation layers. The thickness of each metal conductor layer is 500nm and the thickness of each dielectric layer is 200nm. The final capacitance value is 0.6pF. The parameters of the micro grounding via structure are the same as those of Example 1.

[0149] Example 6:

[0150] The similarities with Example 1 will not be repeated here. The difference is that the piezoelectric ceramic stacked micro-displacement device of the adaptive standing wave and bias control unit is the PI company P-841.80 model, with a linear stroke of 15μm and a resolution of 0.1nm when the driving voltage is 0-100V. The high-precision digital-to-analog converter (AD5761R) and control algorithm are the same as those in Example 1.

[0151] Example 7:

[0152] The similarities with Example 1 will not be repeated here. The difference is that the piezoelectric ceramic stacked micro-displacement device of the adaptive standing wave and bias control unit is the PI company P-841.00 model, with a linear stroke of 15μm and a resolution of 1.0nm when the driving voltage is 0-100V. The high-precision digital-to-analog converter and control algorithm are the same as those in Example 1.

[0153] Example 8:

[0154] The similarities with Example 1 will not be repeated here. The differences are: the mixer assembly protrusion height is 0.03mm, the number of three-dimensional spiral capacitor metal layers is 8, the piezoelectric ceramic micro-displacement device resolution is 0.3nm, and the remaining structural parameters are the same as those in Example 1.

[0155] Example 9:

[0156] The similarities with Example 1 will not be repeated here; the differences are as follows:

[0157] The mixer assembly has a protrusion height of 0.00 mm; the three-dimensional spiral capacitor has 10 metal layers; the piezoelectric ceramic micro-displacement device has a resolution of 0.1 nm; and the remaining structural parameters are consistent with those of Example 1.

[0158] Example 10:

[0159] The similarities with Example 1 will not be repeated here; the differences are as follows:

[0160] The mixer assembly has a protrusion height of 0.05 mm; the three-dimensional spiral capacitor has 6 metal layers; the piezoelectric ceramic micro-displacement device has a resolution of 1.0 nm; and the remaining structural parameters are consistent with those of Example 1.

[0161] Comparative Example 1:

[0162] To verify the technical effect of the present invention, a comparative mixer was constructed as a reference.

[0163] The comparative example uses the same WR-28 waveguide body and MA4E2038-1141T diode as the embodiment, but its mixer assembly adopts the traditional surface mount method, that is, the microstrip substrate with diode and circuit is directly pasted onto the inner wall of the waveguide and covered with a surface mount shield with a height of 0.5mm, forming a raised structure of about 0.75mm inside the waveguide.

[0164] Its cross-coupling compensation capacitor is a standard 0.5pF surface-mount ceramic capacitor, which is connected to the waveguide wall through a grounding via through the microstrip substrate.

[0165] This comparative example is equipped with a manually adjustable short-circuit terminal, which is adjusted to its optimal state and fixed at 25°C before testing. It does not have an adaptive adjustment function. The performance of the comparative example is evaluated under completely identical test conditions and procedures as the example.

[0166] Comparative Example 2:

[0167] The similarities with Example 1 will not be repeated here. The difference is that the annular groove processing is cancelled, the mixer assembly adopts the traditional surface mount method, and the multilayer microstrip substrate is directly pasted to the inner wall of the waveguide. The surface mount cylindrical shield with a coverage height of 0.5mm is used. After assembly, the inner surface of the shield protrudes 0.75mm relative to the original inner wall of the waveguide body. All other structural parameters are the same as those in Example 1.

[0168] Comparative Example 3:

[0169] The similarities with Embodiment 1 will not be repeated here. The difference is that the embedded parasitic compensation network is cancelled, and a standard 0.5pF surface-mount ceramic capacitor is used as the cross-coupling compensation capacitor. It is connected to the waveguide wall through a conventional grounding via through the multilayer microstrip substrate. All other structural parameters are the same as those in Embodiment 1.

[0170] Comparative Example 4:

[0171] The similarities with Example 1 will not be repeated here. The differences are as follows: the adaptive standing wave and bias control unit is cancelled, the movable short-circuit terminal of the waveguide body adopts a manual adjustment structure (calibrated to the optimal position at 25°C before testing and then fixed), the DC bias voltage of the mixer diode is fixed at 2.5V (the initial optimal value of Example 1), and the remaining structural parameters (mixer components, parasitic compensation network, etc.) are the same as those in Example 1.

[0172] The above embodiments are systematically compared with the comparative embodiments in terms of conversion loss, port standing wave ratio and environmental adaptability in the 26.5-40GHz frequency band.

[0173] Table 1. Impact of Mixer Assembly Protrusion Height on Performance

[0174] Performance indicators Test conditions Example 1: Protrusion 0.00mm Example 2: Protrusion 0.03mm Example 3: Protrusion 0.05mm Comparative Example 2: Protrusion 0.75mm Conversion loss (dB) Initial state (25℃) 4.8 5.0 5.3 6.5 After temperature change (65℃) 4.9 5.3 5.7 8.1 LO-IF isolation (dB) Initial state (25℃) 45 43 41 36 After temperature change (65℃) 44 41 38 28 Output the third-order cutoff point (OIP3, dBm). Initial state (25℃) +22.5 +22.0 +21.5 +18.5 After temperature change (65℃) +22.1 +21.3 +20.5 +15.7

[0175] Table 1 clearly demonstrates the significant impact of mixer component bulge height on performance.

[0176] From a theoretical perspective, the distribution of the electromagnetic field inside the waveguide is crucial to mixing efficiency, especially the position of the antinodes of the standing wave field, which determines the maximization of energy coupling. When the height of the mixer component bulge increases, it disrupts the coplanarity of the waveguide inner walls, causing electromagnetic field disturbances and phase mismatch, resulting in increased conversion loss, decreased isolation, and deteriorated linearity.

[0177] The greater the bump height, the more severe the disruption to the standing wave field distribution, causing the mixer diode to deviate from the optimal energy coupling point, thereby reducing conversion efficiency and increasing port crosstalk. As temperature rises, thermal expansion further amplifies this deviation; for example, the conversion loss in Example 3 deteriorated from 5.3dB to 5.7dB, while the performance degradation in Comparative Example 2 was even more severe, with the conversion loss increasing from 6.5dB to 8.1dB. This highlights the importance of coplanar design for high-frequency stability. Therefore, the zero-bump design in Example 1, through the precise integration of the annular groove and the shield, minimizes electromagnetic field disturbances and ensures optimal performance.

[0178] Table 2. Impact of Compensation Capacitor Value and Type on Performance

[0179] Performance indicators Test conditions Example 40.4pF Spiral Capacitor Example 10.5pF Spiral Capacitor Example 50.6pF Spiral Capacitor Comparative example: 30.5pF surface mount capacitor Conversion loss (dB) Initial state (25℃) 5.0 4.8 4.9 5.8 After temperature change (65℃) 5.2 4.9 5.1 6.2 LO-IF isolation (dB) Initial state (25℃) 43 45 44 40 After temperature change (65℃) 41 44 42 36 Output the third-order cutoff point (OIP3, dBm). Initial state (25℃) +21.8 +22.5 +22.0 +19.5 After temperature change (65℃) +21.2 +22.1 +21.4 +17.8

[0180] Table 2 reveals the key role of compensation capacitor value and type in mixer performance.

[0181] In principle, embedded parasitic compensation networks aim to counteract the stray reactance of mixer diodes through a low parasitic inductance structure, while the deviation and type of capacitance value directly affect the compensation effect. The 0.5pF three-dimensional spiral capacitor in Example 1 provides optimal capacitive compensation, maintaining pure capacitive characteristics over a wide bandwidth, thus resulting in the lowest conversion loss and the highest isolation.

[0182] When the capacitance value deviates to 0.4pF or 0.6pF, the matching degree between the compensation network and the diode parasitic reactance decreases, leading to a slight performance degradation, such as the conversion loss increasing to 5.0dB in Example 4. Comparative Example 3 uses surface-mount capacitors, whose long grounding path introduces significant parasitic inductance, triggering self-resonance and causing compensation failure, resulting in a conversion loss as high as 5.8dB and poor temperature stability. This demonstrates that the short-path design of embedded spiral capacitors can effectively suppress high-frequency parasitic effects and ensure consistent performance across the entire frequency band.

[0183] Table 3. Impact of Adaptive Control Resolution on Performance Stability

[0184] Performance indicators Test conditions Example 90.1nm resolution Example 10.3nm resolution Example 71.0nm resolution Comparative Example 4 (Manual Adjustment) Conversion loss (dB) Initial state (25℃) 4.8 4.8 4.8 4.8 After temperature change (65℃) 4.85 4.9 5.1 5.9 LO-IF isolation (dB) Initial state (25℃) 45 45 45 45 After temperature change (65℃) 44.8 44 43 39 Output the third-order cutoff point (OIP3, dBm). Initial state (25℃) +22.5 +22.5 +22.5 +22.5 After temperature change (65℃) +22.3 +22.1 +21.7 +19.1

[0185] The data in Table 3 highlight the impact of adaptive control resolution on performance stability.

[0186] From a principle perspective, the adaptive standing wave ratio (VSWR) and bias control unit compensates for environmental changes by adjusting the position of the movable short-circuit terminal and the bias voltage in real time. High-resolution micro-displacers enable more refined gradient ramping algorithms, quickly locking the peak value of the performance evaluation function, thus minimizing performance fluctuations after temperature changes. Conversely, with lower resolution, insufficient control precision fails to compensate for temperature-induced phase drift, leading to a deterioration in conversion loss to 5.1 dB. The manual adjustment in Comparative Example 4 completely lacks dynamic adaptability, with conversion loss surging to 5.9 dB at 65°C, demonstrating that nanometer-level resolution is crucial for maintaining the environmental robustness of high-frequency mixers, ensuring simultaneous optimization of conversion gain and linearity.

[0187] Table 4. Comparison of overall performance of cross-examples

[0188] Performance indicators Test conditions Example 1: Optimal Combination Example 8, medium combination Example 10 Poor Combination Comparative Example 1: Traditional Structure Conversion loss (dB) Initial state (25℃) 4.8 5.0 5.5 6.2 After temperature change (65℃) 4.9 5.3 6.0 7.8 LO-IF isolation (dB) Initial state (25℃) 45 43 40 38 After temperature change (65℃) 44 41 36 31 Output the third-order cutoff point (OIP3, dBm). Initial state (25℃) +22.5 +21.8 +20.5 +18.0 After temperature change (65℃) +22.1 +21.0 +19.2 +15.2

[0189] Table 4 provides a comprehensive comparison of the performance of different design combinations, explaining the necessity of multi-module collaborative optimization in principle.

[0190] Example 1, as the optimal combination, integrates coplanar mixing components, precise parasitic compensation, and adaptive control, achieving low conversion loss, high isolation, and excellent linearity. Examples 8 and 10 compromise on some parameters, resulting in a gradual decline in performance, demonstrating that defects in any single module can compromise the overall electromagnetic compatibility of the system. The traditional structure in Comparative Example 1, due to surface mount protrusions, surface-mount capacitors, and fixed adjustment, cannot solve the problems of positional deviation and dynamic adaptation, exhibiting the worst performance and low temperature stability.

[0191] This demonstrates that the present invention eliminates high-frequency bottlenecks through integrated design, while the adaptive mechanism dynamically offsets variations in the working environment, achieving optimal global performance.

[0192] Based on the preferred embodiments of the present invention described above, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A radio frequency mixer with integrated adaptive adjustment function, characterized in that, include: The waveguide body defines an electromagnetic propagation channel inside and is provided with a first port for feeding in a first frequency signal, a second port for feeding in a second frequency signal, a third port for deriving a difference frequency signal, and a movable short-circuit terminal. At least two integrated mixer components are symmetrically arranged on the inner wall of the waveguide body with a preset spacing; wherein the preset spacing is set to be an odd multiple of half the wavelength of the first frequency signal inside the waveguide body. An embedded parasitic compensation network is conformally integrated with the integrated mixer assembly within the wall structure of the waveguide body; An adaptive standing wave and bias control unit establishes a control connection with the movable short-circuit terminal of the waveguide body and the integrated mixer assembly; Each of the integrated mixer components includes: An annular groove formed on the inner wall of the waveguide body; A cylindrical shielding cover is embedded and fixed in the annular groove, wherein the inner surface of the cylindrical shielding cover is coplanar with the original inner wall surface of the waveguide body; A multilayer microstrip substrate disposed at the bottom of the annular groove, wherein the top metal pattern of the multilayer microstrip substrate includes mixer diode mounting pads, intermediate frequency output connection lines, and parasitic compensation network microstrip lines; and A mixer diode mounted on the multilayer microstrip substrate, wherein the active junction region of the mixer diode is located in a direction perpendicular to the inner wall of the waveguide body. The center-to-center distance between the two annular grooves is an odd multiple of half the wavelength of the first frequency signal within the waveguide.

2. The RF mixer with integrated adaptive adjustment function according to claim 1, characterized in that: The inner surface of the waveguide body is chemically polished to reduce the surface roughness to Ra0.4μm, and a nickel transition layer and a pure silver conductive layer are deposited sequentially thereon; wherein the thickness of the nickel transition layer is 1-2μm and the thickness of the pure silver conductive layer is 4-5μm.

3. The RF mixer with integrated adaptive adjustment function according to claim 1, characterized in that: The cylindrical shield is made of Kovar alloy and has a gold-plated inner and outer surface; the mixer diode is a gallium arsenide Schottky beam-leaded diode.

4. The RF mixer with integrated adaptive adjustment function according to claim 1, characterized in that: The embedded parasitic compensation network is disposed on the outer wall of the waveguide body and spatially aligned with the annular groove of the integrated mixer assembly. The embedded parasitic compensation network includes: Three-dimensional spiral capacitor structure; And a micro-grounding via penetrating the main wall of the waveguide; In this configuration, one electrode of the three-dimensional spiral capacitor structure is connected to a port of the mixer diode within the integrated mixer assembly through the micro grounding via, while the other electrode is directly connected to the outer wall of the waveguide body, which serves as the system's common ground, to achieve a large-area, low-impedance connection.

5. The RF mixer with integrated adaptive adjustment function according to claim 4, characterized in that: The three-dimensional spiral capacitor structure is formed by multilayer thin film deposition and photolithography, consisting of multiple alternating stacked metal conductor layers and high dielectric constant dielectric layers; the micro grounding via is formed by laser drilling and filled with conductive silver paste, and its physical length is limited to the wall thickness of the waveguide body.

6. The RF mixer with integrated adaptive adjustment function according to claim 4, characterized in that: The micro grounding via is formed by picosecond laser drilling, with a diameter of 0.05-0.1 mm, and is filled with highly conductive silver paste. The highly conductive silver paste consists of 85% silver powder, 10% epoxy resin and 5% reactive diluent, with the silver powder particle size distribution being 1-5 μm.

7. The RF mixer with integrated adaptive adjustment function according to claim 1, characterized in that: The adaptive standing wave and bias control unit includes: An actuator subunit is used to adjust the position of the movable short-circuit terminal and the DC bias voltage of the mixer diode; A sensor subunit is used to monitor the performance indicators of the RF mixer in real time; and The processing and control unit receives the monitoring data from the sensor subunit and outputs control signals to the actuator subunit according to a preset algorithm.

8. The RF mixer with integrated adaptive adjustment function according to claim 7, characterized in that: The actuator subunit includes: A piezoelectric ceramic stack micro-displacement device mechanically coupled to the movable short-circuit end of the waveguide body, the piezoelectric ceramic stack micro-displacement device generating a linear displacement with nanometer-level resolution under the drive of a control signal, so as to fine-tune the position of the movable short-circuit end; and A high-precision digital-to-analog converter is used to provide a precisely adjustable DC bias voltage to the mixer diode according to a control signal.

9. A radio frequency mixer with integrated adaptive adjustment function according to claim 7, characterized in that: The sensor subunit includes: A dual directional coupler is placed near the first port to monitor the forward power and reflected power of the first frequency signal input to the mixer; A power detector is installed at the third port for real-time measurement of the output power of the difference frequency signal; and A signal analysis module for performing spectral analysis on the difference frequency signal and extracting the power values ​​of its third-order intermodulation distortion components.

Images