Interconnect Loss Research Articles

Approaches to Heterogeneous Integration for Millimeter-Wave Applications

Introduction. Enhanced performance of electronic systems can be achieved by heterogeneous integration of different semiconductor technologies. The benefits of heterogeneous integration become obvious when close connections between the devices are provided. The development of integration approaches, enabling functionality and improved performance, appears a relevant task for modern microwave microelectronics.Aim. Review of state-of-the-art and promising heterogeneous integration concepts and techniques in microwave microelectronics.Materials and methods. Eight integration approaches that ensure the connection of devices based on different semiconductor technologies for microwave frequencies are considered: monolithic heterogeneous integration, wafer bonding, micro-transfer printing, embedded chip assembly, print additive manufacturing, wire bonding, flip-chip, and hotvia. The integration approaches are analyzed in terms of their implementation specifics, advantages and disadvantages.Results. Monolithic heterogeneous integration and wafer bonding, as well as micro-transfer printing, despite the minimum interconnections, have a number of fundamental limitations. These limitations are related to the compatibility of various semiconductor technologies and the necessity of high technological capabilities. The technology of embedded chip assembly enables the variability of implementation techniques, which makes it possible to provide unique characteristics, e.g., due to the integration of magnetic materials. However, this approach is associated with a high complexity of integration technological processes. Flip-chip integration ensures minimal interconnect losses due to bump miniaturization. Hot-via, as a modification of flip-chip, provides for a better compatibility with microstrip type circuitry. Their further improvement and mass application largely depends on the development of technologies for the formation of low-pitch interconnections.Conclusion. The development of close integration approaches in microwave microelectronics is proceeding both in the monolithic direction, i.e., monolithic heterogeneous integration wafer bonding, as well as in the quasi-monolithic direction, i.e., micro-transfer printing, embedded chip assembly, print additive manufacturing, flip-chip, and hot-via. The conducted comparative analysis of the presented methods has practical application.

Development of early life gut resistome and mobilome across gestational ages and microbiota-modifying treatments.

Gestational age (GA) and associated level of gastrointestinal tract maturation are major factors driving the initial gut microbiota composition in preterm infants. Besides, compared to term infants, premature infants often receive antibiotics to treat infections and probiotics to restore optimal gut microbiota. How GA, antibiotics, and probiotics modulate the microbiota's core characteristics, gut resistome and mobilome, remains nascent. We analysed metagenomic data from a longitudinal observational study in six Norwegian neonatal intensive care units to describe the bacterial microbiota of infants of varying GA and receiving different treatments. The cohort consisted of probiotic-supplemented and antibiotic-exposed extremely preterm infants (n=29), antibiotic-exposed very preterm (n=25), antibiotic-unexposed very preterm (n=8), and antibiotic-unexposed full-term (n=10) infants. The stool samples were collected on days of life 7, 28, 120, and 365, and DNA extraction was followed by shotgun metagenome sequencing and bioinformatical analysis. The top predictors of microbiota maturation were hospitalisation length and GA. Probiotic administration rendered the gut microbiota and resistome of extremely preterm infants more alike to term infants on day 7 and ameliorated GA-driven loss of microbiota interconnectivity and stability. GA, hospitalisation, and both microbiota-modifying treatments (antibiotics and probiotics) contributed to an elevated carriage of mobile genetic elements in preterm infants compared to term controls. Finally, Escherichia coli was associated with the highest number of antibiotic-resistance genes, followed by Klebsiella pneumoniae and Klebsiella aerogenes. Prolonged hospitalisation, antibiotics, and probiotic intervention contribute to dynamic alterations in resistome and mobilome, gut microbiota characteristics relevant to infection risk. Odd-Berg Group, Northern Norway Regional Health Authority.

Influence of Fe-rich phases and precipitates on the mechanical behaviour of Al-Cu-Mn-Fe-Sc-Zr alloys studied by synchrotron X-ray and neutron

A multiscale methodology using scanning and transmission electron microscope, synchrotron X-ray nano-tomography and micro-tomography, small angle neutron scattering, and in situ synchrotron X-ray diffraction has been used, to reveal the effect of Fe-rich phases and precipitates on the mechanical behaviour of an Al-Cu-Mn-Fe-Sc-Zr alloy. The α-Al grains size is reduced from 185.1 μm (0 MPa) and 114.3 μm (75 MPa) by applied pressure. Moreover, it has been demonstrated that suitable heat treatments modify the 3D morphology of Fe-rich phases from interconnected to a disaggregated structure that improves the mechanical properties of the alloy. The size and morphology evolution of fine precipitates under different ageing temperature and time are revealed. At ageing temperature of 160 °C, the precipitates change from GP zones to θ' (around 75 nm in length) with ageing time increasing from 1 h to 24 h; the Vickers hardness increases from 72.0 HV to 110.7HV. The high ductility of the Sc, Zr modified Al-Cu alloy is related to the complex shape and the loss of interconnectivity of the Fe-rich particles due to the heat treatment. The evolution of the crystal lattice strains in α-Al, and β-Fe calculated during tensile test using in-situ synchrotron X-ray diffraction corroborates the influence of the microstructure in the ductility of the modified alloy.

Preparation of Hydrophobic Low-k Epoxy Resins with High Adhesion Using a Benzocyclobutene-Rosin Modifier

With the development of high-frequency communication, the miniaturization of integrated circuits (ICs) has led to interconnected signal delays and losses. The dielectric properties of epoxy resins are no longer satisfactory, because of the presence of polar groups. Herein, a benzocyclobutene-rosin modifier was designed and synthesized to reduce the dielectric constant of epoxy resins. The recross-linking network of benzocyclobutene inhibits the polarizability of the polar groups, and rosin skeletons increase the free volume of epoxy resins. Consequently, the dielectric constant of epoxy resins decreased from 3.12 to 2.72 and their water absorption decreased from 1.06% to 0.861%. The dielectric constant and water absorption are lower than those of most commercial dielectric materials. Additionally, the adhesion of epoxy resins was improved, which broke the contradiction between the dielectric properties and the adhesion of traditional epoxy resins. The modified epoxy resin was finally applied to printed circuit boards. At 15 GHz, its signal loss was 28.8 dB, which was lower than that of a commercial epoxy resin (FR-4), demonstrating that the modified epoxy resin can better meet the miniaturization needs of ICs.

Implementation of High-Q Embedded Band Pass Filter in Wireless Communication

At 12.8 MHz center frequency, the advanced miniaturized polymer-based planar high quality factor (Q) passive elements embedded bandpass filter works in the L-band. Because most of the demands operate inside the spectrum, the wideband or high-speed operation necessary to enhance must be acquired in microwave frequency ranges. The channel has a quiet, high-performance microfilter with wideband rejection. Capacitors and inductors are used in the high quality factor (Q) passive components, and related networks are incorporated in the filter. Embedded layers are concatenated using Three-Dimensional Integrated Circuit (3D-IC) integration, parasitics are removed, and interconnection losses are negotiated using de-embedding methods. A wireless application-based Liquid Crystalline Polymer (LCP) viewpoint is employed as a substrate material in this work. The polymer processes, their properties, and the incorporated high-Q Band Pass Filter Framework. The suggested filter model is computed and manufactured utilizing the L-band frequency spectrum, decreasing total physical length by 31% while increasing bandwidth by 45%.

Effects of underfill on wideband flip-chip packaging for 5G millimeter-wave applications

This letter investigates the effects of the underfill on the wideband flip-chip packaging for 5G millimeter-wave (mm-Wave) applications. For accurate interconnect design, a new hybrid equivalent circuit model is proposed. Targeting at the phased array systems with high density I/Os, a compact anti-pad structure is implemented and co-designed with the high impedance transmission line and the low-cost 90 μm solder balls, compensating the flip-chip capacitive parasitics and realizing the compact low-loss interconnect. To evaluate the underfill effect on the interconnect parasitics, both theoretical analyses and simulations are undertaken. For demonstration, by using a glass substrate with the fan-out process, back-to-back flip-chip packaging structures are designed, fabricated, and measured. Measured results demonstrate that with and without underfill U8410-99 the interconnect return loss is better than 20 and 10 dB from DC to 90 GHz, with an insertion loss of 0.2 and 0.45dB at 60GHz, respectively.

Understanding Early Failure Behavior in 3D-Interconnects: Empirical Modeling of Broadband Signal Losses in TSV-Enabled Interconnects

We develop an empirical model for measured frequency-dependent insertion loss ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\vert {S}_{{21}}\vert $ </tex-math></inline-formula> ). The model parameters are determined with a stochastic optimization implementation of the Levenberg–Marquard method. We compare measured <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\vert {S}_{{21}}\vert $ </tex-math></inline-formula> on through silicon via (TSV)-interconnects, from two different providers, as a function of the extent of thermal annealing. The frequency-dependent changes in the electrical characteristics of the interconnect are attributed to silanol (Si-OH) and other dangling bond polarizations at the Si–SiO interface between the silicon substrate and the lateral silicon oxide that isolates the coaxial metal core from the silicon substrate. The changes in the polarizations are traceable to changes in the chemistry of the isolation dielectric during thermal annealing. The data also suggest that the evolution of the chemical defects inherent in the “as-manufactured” products may be responsible for some of the signal integrity degradation issues and other early reliability failures observed in TSV-enabled 3-D devices.

A 56-Gb/s PAM-4 Transmitter/Receiver Chipset With Nonlinear FFE for VCSEL-Based Optical Links in 40-nm CMOS

This article presents a 56-Gb/s four-level pulse amplitude modulation (PAM-4) transmitter/receiver (TX/RX) chipset for optical link. The optical transmitter (TX) adopts a digital-to-analog converter (DAC)-based structure with a two-tap nonlinear PAM-4 feed-forward equalizer (FFE) implemented in the digital domain to compensate for the asymmetric responses from the vertical cavity surface-emitting laser (VCSEL). The 56-Gb/s PAM-4 receiver (RX) is connected to the photodiode (PD) and transimpedance amplifier (TIA) module for optical application. The interconnection loss from the RX chip to the PD+TIA module is compensated by the RX continuous-time linear equalizer (CTLE) and the one-tap decision-feedback equalizer (DFE). The pattern-filter-based bang-bang phase detector in the RX improves the jitter performance. Both TX and RX are fabricated in the 40-nm CMOS. The VCSEL TX demonstrates a 56-Gb/s compensated PAM-4 waveform with −0.9-dBm outer optical modulation amplitude (OMA) while consuming only 97 mW of power (1.73 pJ/b). The RX achieves 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−12</sup> bit error rate (BER) with 0.08 UI margin under 10-dB interconnection loss while dissipating 278 mW of power (4.96 pJ/b).

Stories from London's Docklands: Heritage Encounters, Deindustrialization, and the End of Empire

AbstractThis article analyzes the activism of Eastside Community Heritage in London's Docklands, circa 1997 to 2003, following its establishment by community activists concerned by the British National Party's electoral success in the postindustrial area. Eastside attributed local racism to deindustrialization and unaccountable, exclusionary redevelopment. Aiming to recenter solidarity against economic injustice—thereby countering racism—and to challenge redevelopers’ neglect, the group published booklets celebrating the area's working-class past. But the project's archived oral histories show that residents remembered an area forged by different ideals. Their nostalgia was for participation in empire through the docks as much as for an idealized working-class community. Residents rarely distinguished between the interconnected losses of imperial purpose and social cohesion. I frame these tensions as a heritage encounters, making three key arguments. First, memories of youth in the imperial metropole informed residents’ perceptions of the late twentieth-century nation, despite recent scholarly efforts to separate them temporally and conceptually. Second, contrary to their predominance within histories of postwar class identity, deindustrialization and urban change were popularly understood within a larger, postimperial narrative of local and national decline. Finally, a close reading of this project offers a vivid case study into the fragility of British multiculturalism in the 1990s and 2000s.

Semi-additive patterning process based fabrication of miniaturized, package-embedded high conversion ratio inductors for DC-DC converters

• A scalable semi-additive patterning process for fabrication of highly miniaturized toroidal inductors irrespective of bulk conductivity of the magnetic core. • Double sided, 3-step drilling for less tapered through substrate features. • Dielectric slot filling requires higher vacuum time, pressure, and pressure time. • No photoresist seepage into through substrate vias with roll lamination. Many data centers currently operate at low power efficiencies (∼75%) because of the many voltage conversions necessary to step down inputs from 48 V to 1 V. This voltage step-down is accomplished in the Power System-on-Chip (PwrSoC) package, which contain large quantities of surface mount inductors. However, surface mount inductors are large in area and require long power delivery networks to supply the voltage to the PwrSoC, thereby leading to interconnection losses and reducing overall system efficiency. Miniaturizing these inductors could place them nearer to the PwrSoC. Miniaturized and embeddable solenoid and toroidal inductors can be built from magnetic substrates using patterned copper windings created from through substrate vias and micropatterning. However, to achieve inductances close to SMTs, magnetic substrates must be thick or have large lateral footprints. Furthermore, the magnetic flux leakage must be minimized between inductors. This work will elucidate the challenges of dielectric filling of through substrate slots, laser drilling of slots and vias in different substrates, and dry film photoresist lamination that will enable complete copper windings. This paper presents the process flow, challenges, and redressal of these challenges to build miniaturized, embedded inductors that have previously been introduced by our research team.

A W-Band High Radiation Efficiency With BCB-Air Cavity-Backed Antenna Based on Through Silicon Ring Trench

The design of a W-band BCB-air cavity-backed antennas is presented in this letter. The three-dimensional (3-D) capacitance model of the cavity-backed structure is analyzed and the design methodology is illustrated in detail. The proposed antenna is fabricated by using our in-house silicon-based MEMS photosensitive composite film fabrication process by combining localized backside etching technology and through silicon ring trench. The measured impedance bandwidth (−10 dB) of the antenna is from 88.9 to 93.7 GHz with a maximum gain 8.2 dBi and 95% radiation efficiency at 91 GHz, higher the gain and radiation efficiency by 2.2 dB and 35%, respectively, as compared with the case without TSRT. The measured results show close agreement with the simulated results. The proposed antenna can be further 3-D heterogeneously integrated with other components because of low interconnection loss to enhance the high-frequency performance for future millimeter wave radar and communication systems.

A Novel End-Wall Waveguide Excitation With Wide Bandwidth and Simple Structure for Millimeter-Wave/Terahertz Application

In millimeter-wave (MMW)/terahertz (THz) band, the insertion loss of interconnection between waveguide and planar transmission lines is usually high, and the microassembly process is complex. In this letter, a novel broadband end-wall waveguide electromagnetic wave mode excitation is proposed for MMW/THz application. A simple circular open-circuit stub is utilized to transform the quasi-TEM mode of grounded coplanar waveguide (GCPW) to the dominant TE <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">10</sub> mode of rectangular waveguide (RWG). The operating frequency band can be adjusted and improved by controlling the positions of the matching via hole adjacent to RWG port. To further increase the excitation bandwidth, a U-shaped iris (U-iris) on the other side of the substrate is introduced and embedded in the end wall of the RWG to generate dual resonant excitation modes. A 0.2-THz transition from GCPW to RWG is designed as an example. Three back-to-back prototypes with different lengths are fabricated and measured. The measured return loss of single transition is better than 13 dB with an insertion loss of 0.8 ± 0.15 dB in the frequency range of 185–229 GHz (21.2% bandwidth). The proposed end-wall transition has the advantages of high efficiency, easy fabrication, and wideband performance.

Early effects of ovariectomy on bone microstructure, bone turnover markers and mechanical properties in rats

BackgroundFragility fracture is one of the most serious consequences of female aging, which can increase the risk of death. Therefore, paying attention to the pathogenesis of postmenopausal osteoporosis (PMOP) is very important for elderly women.Methods and materialsForty 12-week-old female rats were divided into two groups including the ovariectomy (OVX) group and the control group. Four rats in each group were selected at 1, 4, 8, 12 and 16 weeks after operation. Vertebral bones and femurs were dissected completely for micro-Computed Tomography (micro-CT) scanning, biological modulus detection and histomorphological observation.ResultsIn OVX group, bone volume/total volume (BV/TV), bone trabecular connection density (Conn.D) and trabecular bone number (Tb.N) decreased significantly with time (P < 0.05). The elastic modulus of femur in OVX group was lower than that in control group, but there was no significant difference between them (P > 0.05). Over time, the tartrate resistant acid phosphatase (TRAP), osteocalcin (BGP), type I procollagen amino terminal propeptide (PINP) and type I collagen carboxy terminal peptide (CTX-I) in OVX group increased significantly (P < 0.05). The micrographs of the OVX group showed sparse loss of the trabecular interconnectivity and widening intertrabecular spaces with time.ConclusionThe bone loss patterns of vertebral body and femur were different in the early stage of estrogen deficiency. The bone turnover rate of OVX rats increased, however the changes of biomechanical properties weren’t obvious.

High-density low-loss millimeter-wave package interconnects with the impact of dielectric-material surface roughness

Heterogeneous package integration and chiplet approaches are the key technology to enable next-generation high performance small form-factor packages for emerging applications. Millimeter-wave packaging for fifth-generation and upcoming sixth-generation platforms also need to meet the high-density low signal-loss interconnect specifications utilizing advanced conductor and dielectric materials. This article presents the comparison of the liquid-based photoimageable dielectric (PID) and dry-film dielectric materials in terms of interconnect path losses that are critical in mm-wave frequency bands. The conductor loss being more dominant in the frequency bands and in thinner dielectric structures, we assess daisy chains and microstrip lines on 15-μm dielectric by measuring the S-parameters to quantify the impact of the surface roughness at around 28 GHz. Measured results from the daisy chain and microstrip line structures exhibit that the smooth surface of the liquid-based PID (3 nm) leads to 8%–32% lower signal loss in the dB scale than the 325-nm rough dry-film dielectric. The study provides comprehensive experimental results that the different material forms with various surface roughness largely impact the package-level interconnect loss.

A Hybrid Antenna in Package Solution Using FOWLP Technology

Abstract Widespread millimeter wave applications have promoted rapid development of system in package (SiP) and antenna in package (AiP). Most AiP structures take the form of flip chip onto antenna substrate, where signal interconnect losses are introduced by solder bumps. The integration may be unavailable for chips with fine pad pitches. Fan-out wafer level package (FOWLP) with antenna patterning on redistributed layers (RDL) is another method for millimeter wave AiP. In this project, a hybrid integration AiP structure is developed. The microwave monolithic integrated circuit (MMIC) chip and antenna unit are integrated with chip-first FOWLP process. Multilayer organic substrate with fine pitch RDL interconnections meets the requirements of wideband antenna design. Modified coplanar waveguide is adopted to feed 2 × 2 aperture array formed on RDL layer. Package warpage is measured using Shadow Moire techniques. The antenna forms an aperture-coupled stacked patch array with bandwidth 25% and gain 8.5dBi for 60 GHz digital communication system. The proposed approach is a convenient hybrid integration solution adopting FOWLP process for millimeter wave AiP systems.

FoilMet®‐Interconnect: Busbarless, electrically conductive adhesive‐free, and solder‐free aluminum interconnection for modules with shingled solar cells

AbstractThe investigation of novel cell‐to‐cell interconnection methods has gained importance with the increase of wafer sizes. Shingling (i.e., overlapping) of solar cells is not only a solution for the interconnection of smaller solar cells but also a chance to increase the output power density by (i) increasing the active cell area within the module, (ii) decreasing shading losses, and (iii) reducing electrical interconnection losses. Replacing the commonly used expensive materials such as the silver in the cell's electrodes and the electrically conductive adhesive increases the attractiveness of a shingled interconnection. This article introduces the FoilMet®‐Interconnect, an approach using laser‐welded aluminum foil, for shingling and presents two solutions for configuration. With the production of aluminum foil‐based shingle strings and modules, saving 36.7% of silver, a proof of concept was successful. In subsequent measurements, an efficiency of ηFoilMet = (21.0 ± 0.2)% was achieved with foil‐based strings of three shingled cell. Temperature cycle tests after encapsulation indicate the applicability and durability of the approach.

Reservoir computing based on a silicon microring and time multiplexing for binary and analog operations

Photonic implementations of reservoir computing (RC) promise to reach ultra-high bandwidth of operation with moderate training efforts. Several optoelectronic demonstrations reported state of the art performances for hard tasks as speech recognition, object classification and time series prediction. Scaling these systems in space and time faces challenges in control complexity, size and power demand, which can be relieved by integrated optical solutions. Silicon photonics can be the disruptive technology to achieve this goal. However, the experimental demonstrations have been so far focused on spatially distributed reservoirs, where the massive use of splitters/combiners and the interconnection loss limits the number of nodes. Here, we propose and validate an all optical RC scheme based on a silicon microring (MR) and time multiplexing. The input layer is encoded in the intensity of a pump beam, which is nonlinearly transferred to the free carrier concentration in the MR and imprinted on a secondary probe. We harness the free carrier dynamics to create a chain-like reservoir topology with 50 virtual nodes. We give proof of concept demonstrations of RC by solving two nontrivial tasks: the delayed XOR and the classification of Iris flowers. This forms the basic building block from which larger hybrid spatio-temporal reservoirs with thousands of nodes can be realized with a limited set of resources.

The ultrastructure of cartilage tissue and its swelling response in relation to matrix health.

This multi-length scale anatomical study explores the influence of mild cartilage structural degeneration on the tissue swelling response. While the swelling response of cartilage has been studied extensively, this is the first study to reveal and correlate tissue microstructure and ultrastructure, with the swelling induced cartilage tissue strains. Cartilage sample strips (n=30) were obtained from the distal-lateral quadrant of thirty mildly degenerate bovine patellae and, following excision from the bone, the cartilage strips were allowed to swell freely for 2h in solutions of physiological saline and distilled water successively. The swelling response of this group of samples were compared with that of healthy cartilage, with (n=20) and without the surface layer (n=20). The subsequent curling response of cartilage showed that in healthy tissue it was highly variable, and with the surface removed some samples curved in the opposite direction, while in the mildly degenerate tissue group, virtually all tissue strips curved in a consistent upward manner. A significant difference in strain was observed between healthy samples with surface layer removed and mildly degenerate samples, illustrating how excision of the surface zone from pristine cartilage is insufficient to model the swelling response of tissue which has undergone natural degenerative changes. On average, total tissue thickness increased from 940µm (healthy) to 1079µm (mildly degenerate), however, looking at the zonal strata, surface and transition zone thicknesses both decreased while deep zone thickness increased from healthy to mildly degenerate tissue. Morphologically, changes to the surface zone integrity were correlated with a diminished surface layer which, at the ultrastructural scale, correlated with a decreased fibrillar density. Similarly, fibrosity of the general matrix visible at the microscale was associated with a loss of later interconnectivity resulting in large, aggregated fibril bundles. The microstructural and ultrastructural investigation revealed that the key differences influencing the tissue swelling strain response was (1) the thickness and extent of disruption to the surface layer and (2) the amount of fibrillar network destructuring, highlighting the importance of the collagen and tissue matrix structure in restraining cartilage swelling.

All-Optical Synapse With Directional Coupler Structure Based on Phase Change Material

With the rapid growth of the artificial neural network, the operational efficiency of von Neumann computing architecture is limited by the separation of memory and processor, and the exploration of the efficient hardware mimicking bionic neurons and synapses has become a matter of great urgency. Moreover, the circuit implements for this architecture have potential limitations such as slow switching speed, high computational power consumption, and high interconnection loss. As an alternative, neuromorphic engineering in the photonic domain has recently gained widespread international attention. In this work, we propose an all-optical synaptic device based on a directional coupler structure, which can control the degree of light field distribution variation by changing the state of the phase change material Ge <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> Sb <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> Te <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">5</sub> (GST) distributed in discrete islands on one port. The mode distribution and propagation of the field have been carefully analyzed with pulsed light of different power. More importantly, it is possible to flexibly design the weight update of the synaptic devices by varying the size and location of the GST islands. Verified by the Spiking Neural Network, we can improve the recognition accuracy of handwritten figures from the original 50% to 91% by improving the linearity and accuracy of the synapse weights. This may provide a new solution for future low-power non-volatile photonic integrated circuits.

Single-Ended-to-Balanced Broadband/Filtering Planar Magic-Ts With High Common-Mode Suppression

In this article, a single-ended-to-balanced (SETB) broadband magic-T based on microstrip/slotline transitions and microstrip/slotline T-junctions is proposed. For the purpose of miniaturization and low interconnection loss, an SETB filtering magic-T with a sharp selectivity is also achieved by introducing a pair of microstrip stub-loaded resonators (SLRs) at the two output ports. For the first design, a U-type microstrip feed line integrated with a circular slot stub is adopted at each output port, which makes the differential-mode (DM) response independent of the common-mode (CM) one. Therefore, the design procedure is simplified significantly. Meanwhile, based on the microstrip/slotline T-junctions, in-phase and out-of-phase two-way equal power division networks can be realized, respectively. For the second design, a DM passband with two transmission zeros (TZs) is generated by introducing SLRs based on the first design. By employing the characteristic mode analysis method, it is easy to find that the center frequency and fractional bandwidth (FBW) can be controlled by changing the dimensions of SLR. To the best of the authors' knowledge, the SETB magic-T with broadband or filtering performances may be the first ever reported. To verify the feasibility of the designs, two prototype circuits of the proposed magic-Ts are fabricated and measured. The simulation results are in good agreement with the measured ones. In these two designs, the functions of in-phase and out-of-phase power dividing, broadband or frequency selectivity, isolation between output ports, and CM suppression can be realized at the same time.