Monolithic Integrated Circuit Research Articles

IEEE Topical Meeting on Silicon Monolithic Integrated Circuits in RF Systems Calls for Papers

IEEE Topical Meeting on Silicon Monolithic Integrated Circuits in RF Systems Calls for Papers

Enhanced performance of Ku-band GaN MMIC PA through embedded microfluidic cooling

The impact of embedded microfluidic cooling on the performance of Ku-band gallium nitride (GaN) microwave monolithic integrated circuit (MMIC) power amplifiers (PAs) has been studied in this work. The research demonstrates that embedded cooling can substantially lower junction temperatures, mitigate the self-heating effect, and thereby improve the electrical performance of GaN MMIC PAs. Using merely 10.4 mW of pumping power (Ppump), the technology reduces the maximum junction temperature (Tj,max) from 216.7 to 147.8 °C. Furthermore, the system enhances the saturation current by 13.8%, increases output power (Pout) and gain (Gain) by 1.8 dB, elevates the power added efficiency (PAE) by 5.9%, and improves the heat flux of last gate (Qgate) to 37.3 kW/mm2 and the heat flux of MMIC (QMMIC) over 2.6 W/mm2, while reducing the thermal resistance (Rj-c) from 2.13 to 1.26 °C/W. Additionally, there is a notable improvement of the small signal gain (S21) by 2.5 dB. Therefore, the achievements in this work indicate that embedded cooling offers a powerful technique to suppress the thermal effects on GaN MMIC PAs, consequently improving their electrical performance.

Design and Implementation of a DC to AC Power Electronics-Based Inverter that Produces Pure Sine Wave Output for Critical Engineering Applications

Power inverters play a crucial role in the field of engineering, particularly in applications where power stability is imperative. In devices such as Uninterruptible Power Supplies (UPS), the conversion of raw power to DC, subsequent filtering, and inversion to AC are executed through pure sine wave inverters. These inverters exhibit remarkable stability, making them ideal for powering sensitive equipment like data switches and Remote Terminal Units (RTUs). This study delves into the intricate process of converting DC power into a pristine sine wave signal. The heart of this power conversion lies in the utilization of the KA3525A integrated circuit (IC) in conjunction with MOSFETs of the PN55 series, supported by capacitors and resistors for effective power filtration. The KA3525A, a monolithic IC, encompasses all essential control circuits for a pulse width modulating regulator. Within this IC, a voltage reference, error amplifier, pulse width modulator, oscillator, under-voltage lockout, soft start circuit, and output driver collaborate seamlessly. The MOSFETs function as switches, synchronized with the oscillating signal from the KA3525A IC. This coordination, combined with the filter and other signal conditioning units, enables the conversion process. The design achieves the conversion of raw power into a stable pure sine wave signal of 170V AC at the H-bridge terminals, demonstrating the success of the designed approach.

Modified elliptic response bandpass filtering integrated high‐efficiency GaN MMIC power amplifier at 24–28 GHz

SummaryIn this paper, an out‐of‐band suppression enhanced high‐efficiency millimeter‐wave (mm‐wave) gallium nitride (GaN) microwave monolithic integrated circuit (MMIC) filtering power amplifier (FPA) is presented. A modified elliptic response bandpass filtering (BPF) output matching network (OMN) is used to achieve low loss and high out‐of‐band rejection. The circuit of the presented BPF OMN is given and its synthesized method is introduced. The simulated S‐parameters show that the OMN has a low‐loss in‐band response and a desired out‐of‐band rejection performance with two transmission zeros (TZs). To demonstrate this idea, a two‐stage 24–28‐GHz GaN MMIC PA is designed and implemented in a 0.15‐μm GaN‐on‐SiC process, with a mask area of 2.4 × 1.0 mm2. Experimental results show that the realized MMIC PA achieves a small signal gain of 18.8 dB from 24 to 28 GHz with ±1.2‐dB gain flatness and 40‐dB rejection at 19.43 and 33.8 GHz. The saturated power‐added efficiency (PAE) is higher than 29% from 24 to 28 GHz, with a peak PAE of 36.3% at 27 GHz. The output power is within the range of 33.8 to 35.6 dBm in the operating band. For a 100‐MHz five‐carrier LTE signal, the PA can achieve ACLR better than −40 dBc over the band with digital predistortion.

Study on the single-event burnout mechanism of GaN MMIC power amplifiers

In this Letter, a single-event burnout (SEB) mechanism in gallium nitride (GaN) microwave monolithic integrated circuit power amplifiers with a high linear energy transfer of 78.1 MeV·cm2/mg has been investigated in detail. A typical SEB phenomenon was observed. With the aid of photon emission measurements and scanning electron microscopy, it is found that catastrophic burnout occurs in the power-stage GaN high electron mobility transistors (HEMTs) and the metal–insulator–metal (MIM) capacitors, respectively. For the GaN HEMT, the incident heavy ions will generate electron–hole pairs within it, which can gain enough energy with the transverse high electric field. The high-energy electrons will collide with the lattice near the drain electrode and induce significant electron trapping, which will result in a significant longitudinal local electric field. When a critical electric field is achieved, catastrophic burnout occurs. For the MIM capacitor, the burnout is attributed to the single-event dielectric rupture via severe impact ionization or latent tracks when heavy ions strike it.

A compact and reconfigurable C-band DGS-based power divider using RF MEMS capacitive switch

This article presents a reconfigurable and compact power divider design that utilizes a combination of defect structures in the ground plane (DGS) and capacitance loading using a unique RF MEMS switch on the sapphire substrate. The combination of these two techniques results in a reduced-size reconfigurable power divider, wherein the length of the quarter-wavelength transmission line is reduced to λ/16 from λ/4, and the higher-order harmonics are suppressed up to 40 GHz. The RF MEMS capacitive switch is designed, and its DC characteristics are simulated in coventorware. The RF characteristics of the designed power divider are simulated in Ansys HFSS 19.2. The power divider utilizing RF MEMS technology has been fabricated through surface micromachining, requiring only four masks for fabrication. According to measurement findings, insertion loss and return loss are 3.57 dB and 24.6 dB (at 4.4 GHz) and 3.82 dB and 22.06 dB (at 5.4 GHz) when the switch is on and off, respectively. The total area of the design is 3.1 mm2 (21.3% of the conventional power divider). The proposed design can suppress harmonics up to the 8th order by 10 dB. These features render it a viable option for employment in Radio Frequency Integrated Circuits (RFICs).

Low-temperature process design for inversion mode n-channel thin-film-transistor on polycrystalline Ge formed by solid-phase crystallization

We fabricated an inversion mode n-channel thin-film-transistor (TFT) on polycrystalline (poly-) Ge at low temperatures for monolithic three-dimensional large-scale IC (3D-LSI) and flexible electronics applications. Based on our previously reported solid-phase crystallization (SPC) method, we designed an n-channel TFT fabrication process with phosphorous ion implantation to provide the source/drain (S/D). We succeeded in fabricating an n-channel TFT with typical electrical characteristics on poly-Ge and confirmed its operation mode to be inversion mode. However, the fabrication process included a high temperature (500 °C) step for S/D activation. To reduce the process temperature, we used a metal-induced dopant activation method and successfully reduced the activation temperature to 360 °C. This combination is expected to pave the way for high-performance 3D-LSI and flexible electronic devices based on SPC-Ge.

Monolithic MHz-frame rate digital SiPM-IC with sub-100 ps precision and 70 μm pixel pitch

This paper presents the design and characterization of a monolithic integrated circuit (IC) including digital silicon photomultipliers (dSiPMs) arranged in a 32 × 32 pixel matrix at 70 μm pitch. The IC provides per-quadrant time stamping and hit-map readout, and is fabricated in a standard 150-nm CMOS technology. Each dSiPM pixel consists of four single-photon avalanche diodes (SPADs) sharing a quenching and subsequent processing circuitry and has a fill factor of 30 %. A sub-100 ps precision, 12-bit time-to-digital converter (TDC) provides timestamps per quadrant with an acquisition rate of 3 MHz. Together with the hit map, the total sustained data throughput of the IC amounts to 4 Gbps. Measurements obtained in a dark, temperature-stable environment as well as by using a pulsed laser environment show the full dSiPM-IC functionality. The dark-count rate (DCR) as function of the overvoltage and temperature, the TDC resolution, differential and integral nonlinearity (DNL/INL) as well as the propagation delays across the matrix are presented. With aid of additional peripheral test structures, the main building blocks are characterized and key parameters are presented.

Simplified Emulation of Active Load Modulation for a Millimeter-Wave GaN MMIC Doherty Power Amplifier Design

Based on large-signal simulations, this article proposes a simplified method to emulate the active load modulation between the carrier and peaking transistors in a Doherty power amplifier (DPA) design. In contrast to the conventional load–pull method, the proposed method contributes to emulating the active load modulation between the load-modulated transistors with considering their nonlinear interaction. It offers the advantages of predicting the real-world performance of the load-modulated transistors and providing the optimum load-modulation trajectories to be synthesized in a DPA design. To verify the proposed method, a two-way millimeter-wave (mmWave) gallium nitride (GaN) microwave monolithic integrated circuit (MMIC) DPA is designed from 25.5 to 27 GHz. The measured results show that the realized MMIC DPA achieves power-added efficiency (PAE) higher than 30% at saturation, 27% at 6-dB back-off, and 21.5% at 8-dB back-off, with a saturated output power higher than 31.4 dBm across the band. When tested using a 400-MHz modulated signal with digital predistortion (DPD), the MMIC DPA shows adjacent channel power ratio (ACPR) levels better than <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$-$</tex-math> </inline-formula> 40 dBc over the band.

Construction of All Multilayer Monolithic RSMTs and Its Application to Monolithic 3D IC Routing

Monolithic three-dimensional (M3D) integration allows ultra-thin silicon tier stacking in a single package. The high-density stacking is acquiring interest and is becoming more popular for smaller footprint areas, shorter wirelength, higher performance, and lower power consumption than the conventional planar fabrication technologies. The physical design of M3D integrated circuits requires several design steps, such as three-dimensional (3D) placement, 3D clock-tree synthesis, 3D routing, and 3D optimization. Among these, 3D routing is significantly time consuming due to countless routing blockages. Therefore, 3D routers proposed in the literature insert monolithic interlayer vias (MIVs) and perform tier-by-tier routing in two substeps. In this article, we propose an algorithm to build a routing topology database (DB) used to construct all multilayer monolithic rectilinear Steiner minimum trees on the 3D Hanan grid. To demonstrate the effectiveness of the DB in various applications, we use the DB to construct timing-driven 3D routing topologies and perform congestion-aware global routing on 3D designs. We anticipate that the algorithm and the DB will help 3D routers reduce the runtime of the MIV insertion step and improve the quality of the 3D routing.

Systematic Study and Review of InP‐based Tera‐Hertz‐ICs Fabrication Process Technology for Beyond 5G/6G Wireless Communication Networks

Next‐generation “Beyond 5G (B5G)/6G” wireless network systems have been researched and developed for meeting the rapid growth of mobile traffic. A high‐yield fabrication process of InP‐based transistors and tera‐hertz monolithic integrated circuits with 300 GHz operation is reported, which is a candidate frequency for “Beyond 5G/6G” network systems. The main focus is on a high‐yield and reproductive fabrication process of InP‐based high electron mobility transistors (HEMTs) while revisiting previous studies. The DC and radio frequency characteristics of fabricated InP‐HEMTs from the viewpoint of integrating circuits are described. Then, high‐frequency performances of ft = 280 GHz and fmax = 860 GHz in the bias conditions of Vgs/Vds = 0.2/1.1 V with good uniformity are obtained. Finally, InP‐based 300 GHz‐band mixer and power amplifiers are introduced, to which a backside process is also applied for ensuring their stability and enhancing output power.

(Invited) III-V Integration on Silicon for Resource-Efficient Sensor-Technology

This work deals with the wafer-level integration of advanced group III-V devices and integrated circuits on silicon substrate for RF-sensor integration, such as radar functions a very high frequencies beyond 300 GHz [1]. The aim is to achieve both performance improvements on device level, co-integration with digital functions, and advanced integration to achieve a greener usage of material critical to the environment. Submillimeter-Wave frequency bands beyond 300 GHz allow for broadband transmit and receive windows, serviceable to both communications and radar-based applications—increasing data rates and imaging resolutions, respectively. On the other hand, CMOS co-integration is called for by the data acquisition- and other mixed-mode- and fast digital functions. As examples of the integration schemes Terahertz Monolithic Integrated Circuit amplifiers (TMICs) are implemented in an advanced transferred-substrate InGaAs-channel HEMT technology with 20-nm gate length on silicon. The inverted III-V HEMT heterostructure is grown by molecular beam epitaxy (MBE) on 100-mm semi-isolating GaAs wafers and transferred to silicon substrates by using a SiO2-based wafer bond process with subsequent wafer thinning and removal of the GaAs substrate. Thus, only a 100-nm-thick III-V heterostructure layer is remaining on the Si substrate. This advanced transferred-substrate technology also offers the implementation of HEMT devices with backside gate [2,3] to achieve better sub-threshold slope, or field plates to increase both channel confinement and higher breakdown voltages. The 20-nm InGaAs-OI HEMT technology features typical values for the OFF-state breakdown voltage of 5 V and and maximum drain-current density of 1200 mA/mm, respectively. A maximum transconductance of 2400 mS/mm is achieved. The expected cutoff frequency values fT and fmax are above 500 GHz and 1 THz, respectively [4]. A fully passivated back-end-of-line (BEOL) process is used, including three metal layers (MET1–MET3). A NiCr 50 Ohm sq thin-film-resistor layer, as well as an SiN layer for the implementation of MIM capacitors between MET2 andMET3. S-parameter characteristics of a six-stage and nine-stage TMIC amplifiers in the frequency band from 620 to 730 GHz are given as examples. During the on-wafer characterization, the HEMT devices in cascode configuration have been biased at VD= 2 V (1 V drain–source voltage per device) and a current of 350 mA/mm. The measured small-signal gain of the six-stage cascode TMIC amplifier is in the range of 22–25 dB over the frequency range from 670 to above 700 GHz. This corresponds to 4 dB of gain per cascode stage around the 670-GHz frequency range. A nine-stage TMIC amplifier, on the other hand, achieves at least 30 dB of measured gain from 660 to about 700 GHz. This again corresponds to a gain per stage below 4 dB. Such results prove both the advancements in integration as well as state-of-the-art circuit performance co-integrated on silicon.

An improved harmonics suppression power amplifier circuit using integrated self‐resonant capacitor for radar application

SummaryIn order to ensure that the electromagnetic compatibility and mutual interference of multiple radar network systems meet the requirements, a 55‐W gallium nitride (GaN) high harmonics suppression (HSs) high power amplifier (HPA) microwave monolithic integrated circuit (MMIC) with an operating frequency of 5.1–6.3 GHz is designed for radar network system. An HSs method using self‐resonant capacitor integrated in the output matching network of HPA is proposed to improve the HSs characteristics. The measured results show that under the condition of 100‐μs pulse width and 10% duty cycle, the HSs are greater than 48 dBc while the output power is greater than 55 W, the power gain is 21 dB, and the power additional efficiency (PAE) is 54%, which is suitable for use in radar network system. The size of the HPA is 3.55 × 4.05 mm2.

Fabrication of a Novel GaAs Thermopile Sensor for RF Amplitude Demodulation

In this letter, a novel fabrication of straight-channel RF AM demodulator based on thermopile is presented. Its structure consists mainly of CPW line fed with AM signal, resistors and thermopiles to detect the demodulated signal. The structure is fabricated using GaAs Microwave Monolithic Integrated Circuit (MMIC) technology. Experiments are implemented and the results show that this thermopile-based demodulator is able to perform demodulation of AM signal with carrier frequency of 0.01-10 GHz and signal frequency of under 5 kHz. The influence of thermopile length on the performance is also studied and the result shows that when the thermopile length gets larger, the output amplitude of the demodulated signal gets larger, but its bandwidth gets smaller. In our scenario, the length of the thermopile can be set as 50 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu$</tex-math> </inline-formula> m as a tradeoff between amplitude and bandwidth. 2022-0187

A Ku-Band GaN-on-Si MMIC Power Amplifier with an Asymmetrical Output Combiner.

In this paper, a microwave monolithic integrated circuit (MMIC) high-power amplifier (HPA) for Ku-band active radar applications based on gallium nitride on silicon (GaN-on-Si) is presented. The design is based on a three-stage architecture and was implemented using the D01GH technology provided by OMMIC foundry. Details on the architecture definition and design process to maximize delivered power are provided along with stability and thermal analyses. To optimize the amplifier performance, an asymmetry was included at the output combiner. Experimental results show that the HPA achieves a 39.5 dBm pulsed-mode output power, a peak linear gain of 23 dB, a drain efficiency of 27%, and good input/output matching in the 16-19 GHz frequency range. The chip area is 5 × 3.5 mm2 and for the measurements was mounted on a custom-made module. These results demonstrate that GaN-on-Si-based Solid-State Power Amplifiers (SSPAs) can be used for the implementation of Ku-band active radars.

SOI Technology:An Opportunity for RF Designers?

This last decade silicon-on-insulator (SOI) MOS-FET technology has demonstrated its potentialities for highfrequency (reaching cutoff frequencies close to 500 GHz forn-MOSFETs) and for harsh environments (high temperature,radiation) commercial applications. For RF and system-on-chip applications, SOI also presents the major advantage ofproviding high resistivity substrate capabilities, leading to sub-stantially reduced substrate losses. Substrate resistivity valueshigher than 1 kΩΩΩcm can easily be achieved and high resistivitysilicon (HRS) is commonly foreseen as a promising substratefor radio frequency integrated circuits (RFIC) and mixed sig-nal applications. In this paper, based on several experimentaland simulation results the interest, limitations but also pos-sible future improvements of the SOI MOS technology arepresented

A Ku-Band Broadband Stacked FET Power Amplifier Using 0.15 μm GaAs pHEMT.

To meet the application requirements of broadband radar systems for broadband power amplifiers, a Ku-band broadband power amplifier (PA) microwave monolithic integrated circuit (MMIC) based on a 0.15 µm gallium arsenide (GaAs) high-electron-mobility transistor (HEMT) technology is proposed in this paper. In this design, the advantages of the stacked FET structure in the broadband PA design are illustrated by theoretical derivation. The proposed PA uses a two-stage amplifier structure and a two-way power synthesis structure to achieve high-power gain and high-power design, respectively. The fabricated power amplifier was tested under continuous wave conditions, and the test results showed a peak power of 30.8 dBm at 16 GHz. At 15 to 17.5 GHz, the output power was above 30 dBm with a PAE of more than 32%. The fractional bandwidth of the 3 dB output power was 30%. The chip area was 3.3 × 1.2 mm2 and included input and output test pads.

From Netlist to Manufacturable Layout: An Auto-Layout Algorithm Optimized for Radio Frequency Integrated Circuits

Layout stitching is a repetitive and tedious task of the radio frequency integrated circuit (RFIC) design process. While academic research on layout splicing algorithms mainly focuses on analog and digital circuits, there is still a lack of well-developed algorithms for RFICs. An RFIC system usually has a symmetrical layout, such as transmitter and receiver components, low-noise amplifier (LNA), an SPDT switch, etc. This paper aims to address this gap by proposing an automated procedure for the layout of RFICs by relying on the basic device/PCell structure based on the interconnection among circuit topologies. This approach makes the in-series generation of layouts and automatic splicing based on circuit logic possible, resulting in superior stitching performance compared with related modules in Advanced Design System. To demonstrate the physical application possibilities, we implemented our algorithm on an LNA and a switch circuit.

Silicon-thulium hybrid microdisk lasers with low threshold and wide emission wavelength range.

We demonstrate low-threshold and wide emission wavelength range hybrid-integrated silicon-thulium microdisk lasers based on a pulley-coupled design. The resonators are fabricated on a silicon-on-insulator platform using a standard foundry process and the gain medium is deposited using a straightforward, low-temperature post-processing step. We show lasing in 40- and 60-µm diameter microdisks with up to 2.6 mW double-sided output power and bidirectional slope efficiencies of up to 13.4% with respect to 1620 nm pump power launched to the bus waveguides. We observe thresholds less than 1 mW versus on-chip pump power and both single-mode and multimode laser emission spanning across wavelengths from 1825 to 1939nm. These low threshold lasers with emissions over a > 100 nm range open the door to monolithic silicon photonic integrated circuits with broadband optical gain and highly compact and efficient light sources in the emerging ∼1.8-2.0 µm wavelength band.

Equivalent small-signal model of InP-based HEMTs with accurate radiation effects characterization

In this paper, an effective equivalent modeling technique has been proposed to describe small-signal characteristics of InP-based high electron mobility transistors (HEMTs) after proton radiation, which is composed of an artificial neural network and equivalent-circuit models. Small-signal intrinsic parameters of InP-based HEMTs are extracted from S-parameters before and after 2 MeV proton radiation as modeling objects. The deep learning model of a generative adversarial network has been explored to expand the measured finite data samples. Four feedforward neural networks are incorporated to equivalent-circuit topology to form the equivalent model, which are trained to accurately predict the radiation-induced variations of Cgs, Cgd, Rds, and gm, respectively. The prediction accuracy of the developed equivalent model has been well verified in terms of the broad-band S-parameters under radiation fluence of 1 × 1014 and 5 × 1013 H+/cm2. This equivalent modeling method with characterization of radiation damage effects could provide significant guidance for the aerospace monolithic millimeter-wave integrated circuit design.