Low On-state Resistance Research Articles

(Invited) 2D Non-Volatile RF Switches

Radiofrequency (RF) switches are significant components in wireless telecommunication systems. Micro-electro-mechanical system (MEMS) based switches have been widely discussed to replace the conventional solid-state switches that consume static and dynamic energy from their operations. MEMS switches use a mechanical movement, stimulated by a voltage/current, to manipulate an ohmic or capacitive contact. Although MEMS switches have higher linearity and lower power consumption compared to solid-state counterparts, they have some issues such as contact interface degradation, complex fabrication and packaging processes, and large switching voltages (~10-100 V). Therefore, non-volatile switches are preferred because they need no DC voltage and consume zero-static power. Non-volatile resistance switching (NVRS) devices such as resistance-change or phase-change (PC) material based switches have been discussed for RF switch applications. PC materials show a low resistance at the crystalline state and a high resistance at the amorphous state. Compared to MEMS switches, PC switches do not have hermetic packaging requirements and show easier integration with CMOS technologies. However, the phase transition is achieved through Joule heating and thus, the switches need an integrated heater and have slow switching times due to heat transport. Lately, Pi et al. showed memristive RF switch using a pair of electrochemically asymmetric metal electrodes. While memristive switches have power efficiency advantages and smaller physical size than other counterparts, it is necessary to engineer the location and geometry of the metallic filaments for practical applications. Lately, Non-volatile resistance switching (NVRS) phenomenon was first observed in a variety of monolayer semiconducting TMDs sandwiched between metal electrodes. NVRS represents the resistance modulation between a high-resistance state (HRS, ROFF ) and a low-resistance state (LRS, RON ) and then preserve the state without power consumption. Especially, these memristor effects in atomically thin nanomaterials or atomic sheets are labeled as atomristor. Although other solution-processed multilayer two-dimensional (2D) materials have shown NVRS effect, it was believed that monolayer materials can't have an NVRS effect due to its leakage current. However, Sangwan et al. discovered NVRS effects in monolayer MoS2 lateral device using the resistance modulation of its grain boundaries. Nevertheless, for the practical applications, vertical MIM structure is advantageous for smaller device dimension and denser integration. Toward this end, here, we report zero-static power non-volatile RF switches based on MoS2 atomristor. RF circuits for wide-band system require an RF switch having low ON-state resistance and low OFF-state capacitance to achieve a high figure of merits (FOM) cutoff frequency (fc =1/2πRONCOFF ). Our MoS2 RF switch can be programmed with a voltage as low as 0.7 V and has an ON/OFF conductance ratio up to 104. Interestingly, we found that our devices show the one-dimensional area-invariant RON (<10 Ω), and two-dimensional area-dependent COFF , (~20 fF/μm2), yields a higher fc by reducing device area, a defining advantage over PCM switches. We demonstrate the RF performance of the device up to 50 GHz and report 0.3 dB insertion loss, 20 dB isolation (both at 50 GHz), 10 THz cutoff frequency and power-handling capability. Although various RF switches based on MEMS and PC materials have been broadly reported, our non-volatile low-power RF switches based on MoS2 atomristor present a better solution in respect to the described RF performance, non-volatility, and area-dependency.

Reverse-Blocking Normally-OFF GaN Double-Channel MOS-HEMT With Low Reverse Leakage Current and Low ON-State Resistance

An MOS field plate-protected Schottky-drain (gated Schottky-drain) is successfully integrated on a double-channel AlGaN/GaN MOS-HEMT to provide reverse blocking capability. The leakage suppression MOS field plate is deployed on the etched upper GaN channel layer after a barrier fully recess process, leading to a low reverse OFF-state leakage current of −20 nA/mm (at −100 V). The drain metal is deployed adjacent to the MOS field plate, contacting the upper MOS-channel and lower heterojunction channel from the sidewall. A metal-2DEG Schottky contact with a low turn-ON voltage of 0.5 V is achieved. Since the lower channel (below the MOS field plate) is separated from the etched surface of upper GaN channel layer, a high-conductivity MOS-gated channel with a sheet resistance of 806 $\Omega$ /Square is obtained. The device exhibits a threshold voltage of +0.6 V (at $10~\mu \text{A}$ /mm and +1.9 V from linear extrapolation) and an ON-resistance of ~18 $\Omega {\cdot} \text {mm}$ . Besides, a high forward (and reverse) breakdown voltage of 790 V (and −656 V, all at $10~\mu \text{A}$ /mm) is achieved.

A Voltage Multiplier Based High Voltage Gain Transformerless Buck–Boost DC–DC Converter

In this paper, a new transformerless buck–boost converter is introduced. The proposed converter voltage gain is higher that of the conventional buck–boost converter. In the presented converter, only one power switch is used. The switch voltage stress is low, therefore, the low on-state resistance of the power switch can be selected to decrease losses of the switch. The presented converter topology is simple, hence the control of the converter will be simple. The mathematical analyses and principle of the proposed converter are explained. The validity of the proposed converter is confirmed by the experimental results.

AlGaN/GaN High Electron Mobility Transistors with a p-GaN Backgate Structure

GaN based materials are widely used for power devices. AlGaN/GaN heterostructure is able to form high mobility 2DEG channel by its spontaneous and piezoelectric polarization at the AlGaN/GaN interface. Therefore, AlGaN/GaN HEMT is a promising candidate for next generation power device because of its high breakdown voltage, high current density and low on-state resistance (Ron). For efficient power switching applications, the E-mode operation with positive threshold voltage and low leakage current are required. Many efforts have been devoted to make AlGaN/GaN HEMT to achieve E-mode operation, such as gate recess, F-ion implements and p-GaN gate structures. In this study, a p-GaN backgate layer is applied to AlGaN/GaN HEMTs. Additional control ability under the channel layer is observed by applying bias at the p-GaN layer. The device can be turned off easily due to the p-GaN layer under channel layer, which depletes carriers in channel under suitable bias. Simulation and measurement results show the shift of threshold voltage toward positive for possible E-mode operation when backgate bias is negative. The simulation of AlGaN/GaN HEMT with a p-GaN backgate layer was performed with Silvaco TCAD tools including device structures, band diagrams, electron concentrations and ID-VG characteristic. A 30 nm p-GaN backgate layer was implemented under the 30-nm GaN channel with p-type doping of 2 ´ 1017 cm-3. The AlGaN/GaN HEMT epitaxial wafers on low-resistivity Si (111) substrates were grown by MOCVD. Device fabrication started with p-GaN activation in nitrogen atmosphere at 700°C for 15 minutes. Mesa isolation was etched by ICP down to the buffer. The ohmic contact was made by the deposition of Ti/Al/Ni/Au (25/125/45/55 nm) and annealing at 875°C in a nitrogen atmosphere for 40 s. The contact to p-GaN layer was etched by ICP. Backgate metal was made by deposition of Ni/Au (20/20 nm) and annealing at 500°C in an oxygen atmosphere for 5 minutes. Ni/Ti/Al/Ti/Au (30/25/250/25/200 nm) metals were used to form front gate (FG). All the devices under test are with a gate-source distance of 4 μm, a gate length of 2 μm, and a gate-drain distance of 10 μm. The p-GaN backagte layer with bias is able to adjust the fermi levels in the quantum well thus the 2DEG concentration. Therefore, the threshold voltage shifts to positive values while applying negative backgate bias. Electrons are localized in the GaN channel layer when negative backgate bias was applied. This result is to justify the reduction of leakage current because leakage path through buffer layer is confined. From the simulated transfer characteristics of AlGaN/GaN HEMT with a p-GaN backgate under different backgate bias (VBG ). It is clearly to observe that the positive shift of threshold voltage (VTH) is due to the negative VBG . The simulated results showed the positive shift of VTH which was 5 V from VBG of 0 V to -14 V. Therefore, it possible to operate HEMT in either D-mode or E-mode operation by adjusting VBG . The measurement results showed the positive shift of VTH which was 0.55 V from VBG of 0 V to -14 V, with associated off-state leakage current reduction of 73.1%. However the shift of VTH is not as significant as in the simulation, the major reason is due to the bad p-GaN contact. The impact of p-GaN backgate structure in this study on the DC characteristics of AlGaN/GaN HEMTs was investigated. AlGaN/GaN HEMTs with a p-GaN backgate shows an additional control capability in 2DEG channel. Therefore, threshold voltage is adjustable and leakage current is improved by backgate bias. This extra gate control under channel layer is useful for circuit flexibility to achieve both D-mode and E-mode operations in the power device circuit design.

A new high-κ Al2O3 based metal-insulator-metal antifuse

In this paper, a new metal–insulator-metal (MIM) antifuse was fabricated with the high κ Al2O3 deposited by atomic layer deposition (ALD) as the dielectric. On this high κ antifuse structure, the very low on-state resistance was obtained under certain programming conditions. It is the first time that the antifuse on-state resistance has been found decreasing along with the increase of dielectric film thickness, which is attributed to a large current overshoot during breakdown. For the device with a dielectric thickness of 12 nm, very large overshoot current (∼60 mA) was observed and extremely low on-state resistance (∼10 Ω) was achieved.

A Gate Driver of SiC MOSFET for Suppressing the Negative Voltage Spikes in a Bridge Circuit

SiC mosfet has low on-state resistance and can work on high switching frequency, high voltage, and some other tough conditions with less temperature drift, which could provide the significant improvement of power density in power converters. However, for the bridge circuit in an actual converter, high dv/dt during fast switching transient of one mosfet will amplify the negative influence of parasitic components and produce the significant negative voltage spikes on the complementary mosfet , which will threaten its safe operation. This paper proposes a new gate driver circuit for SiC mosfet to attenuate the negative voltage spikes in a bridge circuit. The proposed gate driver adopts a simple voltage dividing circuit to generate a negative gate-source voltage as traditional and a passive triggered transistor with a series-connected capacitor to suppress the negative voltage spikes, which could satisfy the stringent requirements of fast switching SiC mosfet s under the high dc voltage condition with low cost and less complexity. An analysis is presented in this paper based on the simulation and experimental results with the performance comparison evaluated.

Active‐clamped ZVS current‐fed push–pull isolated dc/dc converter for renewable energy conversion applications

This study introduces a new active-clamped current-fed push-pull dc-dc converter for renewable energy conversion applications. The proposed converter conserves small input current pulsation, high-voltage conversion ratio, zero-voltage switching (ZVS) and zero-current-switching operation for primary switches and rectifier diodes, respectively, over a large load range. Also, the additional active clamping circuit serves to suppress the voltage spike across all primary switches, as well as assists in achieving ZVS operation. So, the low-voltage switches with low on-state resistance can be adopted. These features can reduce switching loss, voltage stress, transformer turns ratio and diode reverse-recovery effect. More importantly, similar characteristics can be obtained with minimised switches compared with the reported converters that enable to reduce the cost and improve the reliability. Detailed steady-state operation, analysis, performance comparison, experimental results and loss breakdown of the proposed converter are fully discussed in this study. Finally, a 520 W prototype verifies the theoretical analysis.

A High Reliable High-κ Antifuse Programmed by Intrinsic Overshoot Current

In this paper, we fabricated a W/Al2O3/Ti metal-insulator-metal (MIM) antifuse one-time programmable memory. A large current overshoot was observed during the dielectric breakdown, which generated a strong conductive filament and resulted in extremely low on-state resistance. The performance of this Al2O3-based antifuse has been demonstrated with an ultra large ON/OFF window on the order of 1012. Both breakdown voltage and on-state resistance exhibit excellent uniformity, the off-state lifetime is projected to be 1591 years at 2 V, and the on-state exhibits a reliable filament that is difficult to rupture.

Achieving Efficiencies Exceeding 99% in a Super-Junction 5-kW DC–DC Converter Power Stage Through the Use of an Energy Recovery Snubber and Dead-Time Optimization

A highly efficient 5-kW bidirectional dc–dc converter power stage operating from a 400-V supply implementing super-junction (SJ) metal–oxide–semiconductor field-effect transistor (MOSFETs) is presented. SJ MOSFETs have low on-state resistances and low switching losses. However, their application in voltage-source converters can be compromised by the reverse recovery behavior of their intrinsic diodes and their highly nonlinear output capacitances. A series switching-aid circuit is used to control the output capacitance charging current. The dead times between switching transitions are assessed and optimized in order to deactivate the intrinsic diodes. The combination of these techniques enables very high efficiencies to be attained. Calorimetric measurements indicate a full-load efficiency of 99.1% for the prototype 5-kW dc–dc converter power stage. A loss reduction of approximately 50% is achieved with the prototype converter power stage when compared to an equivalent insulated-gate bipolar transistor (IGBT)-based power stage. Lastly, a loss versus duty cycle function is experimentally determined, which can be used to inform the design of a maximum efficiency point tracking system.

High conversion ratio DC-DC converter with isolated transformer and switched-clamp capacitor for Taiwan photon source

A new high step-up voltage converter that combines a switch capacitor and isolated transformer, together with a passive clamp circuit, is employed to reduce voltage stress on the main power switch. The voltage stress of the power switch should be clamped to 1/4 Vo, and the proposed converter can achieve high step-up voltage gain with appropriate duty ratio. The energy of the leakage inductor can be recycled by the clamp capacitor because of the passive clamp circuit, and low On-state resistance RDS(on) of the power switch can be adopted to reduce the conduction loss. In this paper, several mathematical derivations are presented, CCM and DCM operating principle are discussed, and experimental results are provided to verify the effectiveness of converter topology. Finally, a 24-V-input voltage to 200-V-output voltage and a 150 W output power prototype converter are fabricated in the laboratory.

Characterization of high κ Al2O3 based metal-insulator-metal antifuse

A new metal-insulator-metal (MIM) antifuse structure was fabricated with the amorphous Al2O3 synthesized by atomic layer deposition (ALD) as the dielectric and with either Ti or TiN as top electrode. The device with Ti electrode exhibits lower breakdown voltage compared to the TiN counterpart due to the more oxygen vacancies in Ti electrode antifuse. During breakdown, the effective programming current is found to be several orders larger than the compliance current, and resulting in extremely low on-state resistance. A hypothesis based on the electron trap of oxygen vacancy in the bulk of Al2O3 is proposed to explain this breakdown characteristics.

A Reconfigurable Dual-Frequency Narrowband CMOS LNA Using Phase-Change RF Switches

This paper presents a dual-band low-noise amplifier (LNA) that can be reversibly configured between 3 and 5 GHz using a phase-change (PC) RF switch. Simultaneous noise and conjugate input matching is achieved at the two frequencies using a single PC RF switch and coupled source inductors, thereby minimizing integration complexity and enabling a compact design roughly equal in size to a single-band design. The PC switch has extremely low on-state resistance (1.5– $4\Omega$ ) and parasitic capacitance (12–16 fF), and has a compact footprint ( $\sim 50~\mu \text{m}^{2}$ ). A chip containing the PC switch is fabricated in-house while the rest of the LNA is fabricated in a 0.13- $\mu \text{m}$ CMOS technology. The two chips are combined using an in-house flip-chip integration process. Characterization from five prototypes fabricated in two batches is presented. The integrated LNAs achieved peak gain greater than 20 dB and minimum noise figure less than 2.9 dB in both the frequency bands while consuming a peak power of 7.2 mW from a 1.2-V supply. Comparison of the CMOS-PC prototypes with carefully designed control LNAs demonstrates the effectiveness of the PC switch for RF reconfiguration, which allows the reconfigurable LNA to achieve performance comparable to its nonreconfigurable counterparts.

Implementation of an Active-Clamped Current-Fed Push–Pull Converter Employing Parallel-Inductor to Extend ZVS Range for Fuel Cell Application

This paper proposes a wide-range zero-voltage-switching (ZVS) active-clamped current-fed push–pull converter for fuel cell application. The presented converter achieves ZVS for all of the primary switches and ZCS for secondary diodes from rated load to 10% full load over the wide input voltage variation, improving the overall efficiency. The ZVS of switches is realized by the energy stored in the transformer leakage inductance aided by the secondary parallel inductor. An additional active clamping circuit suppresses the voltage spike as well as assists in realizing ZVS of switches. Low-voltage-rated switches with low on-state resistance can be used. Moreover, the voltage-doubler rectifier is adopted to release the reverse-recovery problem of rectifier diodes and reduce turns ratio of the high-frequency transformer. Detailed steady-state operation, analysis, design, control, comparative study, and experimental results are discussed in depth in this paper. Finally, a 200-W prototype has been built in the laboratory to verify the effectiveness of the proposed converter.

High Breakdown Voltage (−201) $\beta $ -Ga2O3 Schottky Rectifiers

$\beta $ -Ga2O3 Schottky barrier diodes were fabricated in a vertical geometry structure consisting of Ni/Au rectifying contacts without edge termination on Si-doped epitaxial layers ( $10~\mu \text{m}$ , $\text{n}\sim 4\times 10^{15}$ cm $^{-3})$ on Sn-doped bulk Ga2O3 substrates with full-area Ti/Au back Ohmic contacts. The reverse breakdown voltage, ${V} _ \text {BR}$ , was a function of rectifying contact area, ranging from 1600 V at $3.1\times 10^{-6}$ cm2 (20- $\mu \text{m}$ diameter) to ~250 V at $2.2{\times }10^{-3}$ cm $^{-2}$ (0.53-mm diameter). The current density near breakdown was not strongly dependent on contact circumference but did scale with contact area, indicating that the bulk current contribution was dominant. The lowest ON-state resistance, ${R} _\text {on}$ , was 1.6 $\text{m}\Omega \cdot $ cm2 for the largest diode and 25 $\text{m}\Omega \cdot $ cm2 for the 1600-V rectifier, leading to a Baliga figure-of-merit ( ${V} _\text {BR}^{2}/{R} _\text {on})$ for the latter of approximately 102.4 MW $\cdot $ cm $^{-2}$ . The ON-OFF ratio was measured at a forward voltage of 1.3 V and ranged from $3\times 10^{7}$ to $2.5\times 10^{6}$ for reverse biases from −5 to −40 V and showed only a small dependence on temperature in the range from 25 °C to 100 °C.

Investigation of enhancement mode HfO2 insulated N-polarity GaN/InN/GaN/In0.9Al0.1N heterostructure MISHEMT for high-frequency applications

In this paper, we examined normally-OFF N-polar InN-channel Metal insulated semiconductor high-electron mobility transistors (MISHEMTs) device with a relaxed In0.9Al0.1N buffer layer. In addition, the enhancement-mode operation of the N-polar structure was investigated. The effect of scaling in N-polar MISHEMT, such as the dielectric and the channel thickness, alter the electrical behavior of the device. We have achieved a maximum drain current of 1.17A/mm, threshold voltage (VT) =0.728V, transconductance (gm) of 2.9Smm−1, high ION/IOFF current ratio of 3.23×103, lowest ON-state resistance (RON) of 0.41Ωmm and an intrinsic delay time (τ) of 1.456 Fs along with high-frequency performance with ft/ fmaxof 90GHz/109GHz and 180GHz/260GHz for TCH =0.5nm at Vds =0.5V and 1.0V. The numerically simulated results of highly confined GaN/InN/GaN/In0.9Al0.1N heterostructure MISHEMT exhibits outstanding potential as one of the possibility to replace presently used N-polar MISHEMTs for delivering high power density and frequency at RF/power amplifier applications.

Adjoint Method for Increasing the Breakdown Voltage and Reducing the On-State Resistance in Wide Band Gap Power Transistors

A central problem in the power semiconductor industry is the design of wide band gap (WBG) devices with high breakdown voltage (BV) and low on-state resistance (RON). This problem is particularly difficult to solve because high BV requires large devices, in which the separation between the source and drain electrodes in the case of FETs or anode and cathode in the case of p-n junctions is relatively large, while low RON requires relatively small devices with high doping concentrations. For any given semiconductor material there is a trade-off between BV and RON, trade-off that is generally given by the overall dimensions and geometry of the device. In this work we focus on the optimization of standard WBG MOSFET configurations [1], in which we optimize the doping profiles in order to increase the BV while keeping the on-state resistance constant. Our approach is based on applying an adjoint space technique that our group has recently developed for the design of nanoscale semiconductor devices, and which can be used to perform a multidimensional optimization of the doping concentration at each location inside the device using a doping sensitivity approach [2-4]. Note that traditional optimization techniques based on heuristic methods such as genetic algorithms or evolutionary methods are impractical in this case because of the large number of independent variables. For instance, for a finite element mesh with 104 nodes there are 104 independent variables (given by the doping concentration at each finite element), which makes the problem difficult to solve even on parallel or distributed computer clusters. The method that we introduce in this presentation is based on the computation of the doping sensitivity functions of the BV and RON, which show how sensitive the BV and RON are when we add one additional acceptor or donor impurity inside the semiconductor. The doping sensitivity functions are instrumental in the optimization of WBG power devices. As we will describe in the full paper, the doping sensitivity functions can be computed efficiently by using the adjoint method. They can be coupled with gradient-based optimization methods to compute the optimum doping profiles for acceptors, Na(r), and donors, Nd(r). At the conference we will describe the numerical implementation of the adjoint method for the optimization of WBG power devices and present sample simulation results for a few standard structures. Our initial results show that by properly designing the doping profiles in standard power transistors we can increase the BV by 15-20% without affecting the on-state resistance.

A Case Study on Common Mode Electromagnetic Interference Characteristics of GaN HEMT and Si MOSFET Power Converters for EV/HEVs

Wide bandgap semiconductors, such as gallium nitride (GaN)-based power devices have become increasingly popular in the automotive industry due to their low on-state resistance and fast switching capabilities. These devices are sought to replace silicon (Si) devices in power electronics converters for vehicular applications. GaN devices dissipate less energy during each switching event, thus, GaN converter designers can significantly increase the switching frequency without increasing switching losses, relative to Si converters. However, one item of concern is that the high $dv/dt$ of GaN devices, due to the increased switching speed, has the potential to deteriorate the electromagnetic interference (EMI) emission of power converters, and thus may fail the corresponding regulations for vehicles. To understand these effects, this paper presents an experimental case study on the switching characteristics and common mode (CM) noise generation of a GaN-based half-bridge configuration operating in the synchronous boost mode. A comprehensive comparison has been made between the chosen GaN high electron mobility transistors (HEMTs) and Si MOSFET in terms of switching voltage waveforms, ground leakage currents, and CM noise spectrum. By parametrically increasing the gate resistance of GaN HEMT and Si MOSFET, the tradeoff between converter efficiency and CM noise generation is also quantified.

A High-Speed Inkjet-Printed Microelectromechanical Relay With a Mechanically Enhanced Double-Clamped Channel-Beam

We report a high-speed inkjet-printed three-terminal microelectromechanical (MEM) relay with a double-clamped beam that exploits the enhanced stiffness of the double-clamped structure to improve electrical performance. To minimize mechanical delay and pull-in voltage, the contact gap between the channel-beam and drain, and the stiffness of the beam and shape of the drain was carefully designed and optimized through a 3-D finite element simulation. The double-clamped beam prevents stiction-related failure between the channel-beam and drain despite the contact gap being only 370 nm for a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$&gt; 500~\mu \text{m}$ </tex-math></inline-formula> long beam. The resulting printed relay delivers a turn-ON delay of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$8~\mu \text{s}$ </tex-math></inline-formula> at a gate voltage of 10 V, a pull-in voltage of only 7.2 V, immeasurable off-leakage, excellent subthreshold swing, and a small hysteresis window of 2 V without any bending or collapsing of the beam. The device also shows reliable operation over <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$10^{5}$ </tex-math></inline-formula> cycles while maintaining a high ON/OFF ratio of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$10^{8}$ </tex-math></inline-formula> , and extremely low ON-state resistance of 3.7 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\Omega $ </tex-math></inline-formula> . [2016-0180]

Investigation of Gallium Nitride Devices in High-Frequency LLC Resonant Converters

Newly emerged gallium nitride (GaN) devices feature ultrafast switching speed and low on-state resistance that potentially provide significant improvements for power converters. This paper investigates the benefits of GaN devices in an LLC resonant converter and quantitatively evaluates GaN devices’ capabilities to improve converter efficiency. First, the relationship of device and converter design parameters to the device loss is established based on an analytical model of LLC resonant converter operating at the resonance. Due to the low effective output capacitance of GaN devices, the GaN-based design demonstrates about 50% device loss reduction compared with the Si-based design. Second, a new perspective on the extra transformer winding loss due to the asymmetrical primary-side and secondary-side current is proposed. The device and design parameters are tied to the winding loss based on the winding loss model in the finite element analysis (FEA) simulation. Compared with the Si-based design, the winding loss is reduced by 18% in the GaN-based design. Finally, in order to verify the GaN device benefits experimentally, 400- to 12-V, 300-W, 1-MHz GaN-based and Si-based LLC resonant converter prototypes are built and tested. One percent efficiency improvement, which is 24.8% loss reduction, is achieved in the GaN-based converter.

A Novel Structure for Single-Switch Nonisolated Transformerless Buck–Boost DC–DC Converter

A novel transformerless buck–boost dc–dc converter is proposed in this paper. The presented converter voltage gain is higher than that of the conventional boost, buck–boost, CUK, SEPIC, and ZETA converters, and high voltage can be obtained with a suitable duty cycle. In this converter, only one power switch is utilized. The voltage stress across the power switch is low. Hence, the low on-state resistance of the power switch can be selected to decrease conduction loss of the switch and improve efficiency. The presented converter has simple structure, therefore, the control of the proposed converter will be easy. The principle of operation and the mathematical analyses of the proposed converter are explained. The validity of the presented converter is verified by the experimental results.