Liyan Zhu ; Kevin Bai ; Alan Brown ; Matt McAmmond
Wide-bandgap (WBG) devices are considered to be a better alternative to silicon switches to realize high-efficiency and high-power-density power electronics converters, such as electric vehicle on-board chargers. The two major challenges of GaN devices remain are their relatively high cost ( 5 times as compared to Si) and much smaller footprint than Si, which though is preferred in the high-power density application is preferred but brings thermal challenges. Much like SiC is paralleled with Si, GaN could be paralleled with Si to resolve these challenges. In this paper, GaN HEMTs are paralleled to various Si MOSFETs. Two different triggering approaches are considered, one adds a time delay between gate signals and the other uses a pulse triggering technique. Both methods ensure the GaN endure the switching loss while the Si switches conduct the majority of the current thereby maximizing the advantages of both types of switches. To follow is a comprehensive study of the critical transient processes, such as the gate cross talking between Si and GaN, current commutation in the dead band, voltage spikes during the turn-off caused by parasitics, the thermal performance and the cost analysis. Demonstrated success testing this approach at 400V/80A provides evidence that this is a possible approach in the on-board EV (electric vehicle) battery charger applications. The success of testing under 400V/80A makes it possible to an on-board EV battery charger.
As two exemplary candidates of wide-bandgap devices, SiC MOSFETs and GaN HEMTs are regarded as successors of Si devices in medium-to-high-voltage (>1200 V) and low-voltage (<650 V) domains, respectively, thanks to their excellent switching performance and thermal capability. With the introduction of 650 V SiC MOSFETs and GaN HEMTs, the two technologies are in direct competition in <650 V domains, such as Level 2 battery chargers for electric vehicles (EVs). This study applies 650 V SiC and GaN to two 240 VAC/7.2 kW EV battery chargers, respectively, aiming to provide a head-to-head comparison of these two devices in terms of overall efficiency, power density, thermal performance, and cost. The charger essentially is an indirect matrix converter with a dual-active-bridge stage handling the power factor correction and power delivery simultaneously. These two chargers utilize the same control strategy, varying the phase-shift and switching frequency to cover the wide input range (80–260 VAC) and wide output range (200 V–450 VDC). Experimental results indicated that at the same efficiency level, the GaN charger is smaller, more efficient and cheaper, while the SiC charger has a better thermal performance.
In this paper, an enhancement-mode GaN highelectron mobility transistor (HEMT)-based 7.2-kW single-phase charger was built. Connecting three such single-phase modules to the three-phase grid, respectively, generates a three-phase ~22-kW charger with the> 97% efficiency and > 3.3 – kW/L power density, superior to present Si-device-based chargers. In addition to GaN HEMTs with fast-switching transitions yielding high efficiency, the proposed charger employs the dc/dc stage to control the power factor and power delivery simultaneously, yielding little dc-bus capacitance and thereby high power density. To secure the soft switching for all switches within full voltage and power ranges, a variable switching frequency control with dual phase shifts was adopted at high power, and a triple phase shift was employed to improve the power factor at low power. Both control strategies accommodated the wide input range (80-260 VAC) and output range (200-450 VDC). A closed-loop control for the three-phase charger was realized to minimize the output current ripple and balance the power among three single-phase modules. Experimental results validated this design.
An ac/dc + dual active bridge (DAB) circuit was found as one solution for the high-efficiency and high-power-density electric vehicle charger. One control option is to let the ac/dc part only convert the grid voltage to a double-line-frequency folded sine wave and let the DAB stage handle both the power factor (PF) and power delivery. While conventional single-phase-shift tends to lose zero-voltage switching (ZVS) at light load and the variable-switching-frequency dual-phase-shift (DPS) sacrifices the light-load performance, this letter proposes a multiple-phase-shift control, which allows for a fixed-switching-frequency triple-phase-shift (TPS) control at the light load to enhance the grid power quality. At medium- and heavy-load conditions, a phase-shift jump from TPS to DPS is performed to reduce the circulating current and improve efficiency. The proposed control strategy secures ZVS, realizes unity PF accurately and minimizes the control complexity. Experimental results on a SiC-based 7.2-kW charger validated its effectiveness and the smooth transition between the heavy load and light load.
Most of the present electric vehicle (EV) on-board chargers utilize a conventional design, i.e., a boost-type Power Factor Correction (PFC) controller followed by an isolated DC/DC converter. Such design usually yields a ~94% wall-to-battery efficiency and 2~3kW/L power density at most, which makes a high-power charger, e.g., 20kW module difficult to fit in the vehicle. As described in this paper, first, an E-mode GaN HEMT based 7.2kW single-phase charger was built. Connecting three such modules to the three-phase grid allows a three-phase >20kW charger to be built, which compared to the conventional three-phase charger, saves the bulky DC-bus capacitor by using the indirect matrix converter topology. To push the efficiency and power density to the limit, comprehensive optimization is processed to optimize the single-phase module through incorporating the GaN HEMT switching performance and securing its zero-voltage switching. Close-loop control is implemented to minimize the output current ripple and balance the power among three single-phase modules. Simulation and experimental results validated that the proposed charger has an efficiency of ~98%, and a power density of ~5kW/L@20kW.
As two exemplary candidates of wide-bandgap devices, SiC MOSFETs and GaN HEMTs are regarded as successors of Si devices in medium-to-high-voltage (>1200V) and low-voltage (<650V) domains, respectively, thanks to their excellent switching performance and thermal capability. With 650V SiC MOSFETs coming into being the direct competition of SiC and GaN in <;650V domains is inevitable, such as Level-2 battery chargers for electric vehicles. This paper applies 650V SiC and GaN to two 240VAC/7.2kW EV battery chargers, respectively, aiming to provide a head-to-head comparison of these two devices in terms of the efficiency, power density, thermal and cost, with the same control strategy of varying the phase-shift and switching frequency to cover the wide input range (80VAC~260VAC) and wide output range (200V~450VDC).
An indirect matrix converter is employed directly converting the grid ac to the battery voltage, with the dual-active-bridge taking care of the power factor correction and power delivery simultaneously. Such circuit is regarded as one candidate of the high-efficiency and high-power-density electric vehicle onboard chargers, if the double-frequency current ripple to the battery is tolerated. Instead of optimizing the overall charger, this paper is focused on adopting variable switching frequency with multiple phase shifts to accommodate the wide input range (80-260 Vac) and output range (200 V-450 Vdc). In addition to the phase shift between the transformer primary-side and secondary-side voltage, one extra phase shift is added to the primary-side H-bridge when the instantaneous input voltage is higher than the reflected output, otherwise, to the secondary side. The goal is to secure zero-voltage-switching for all switches at all voltage range. Such control strategy is further optimized incorporating with the switch parasitic capacitance and deadband settings. To further enhance the charger performance, GaN HEMTs are equipped to the on-board charger aiming at higher efficiency and higher power density than Si devices. Experimental results indicated that such charger with proposed control strategy embraces the peak efficiency of >97% at 7.2 kW and a power density of ~4 kW/L.
At present time, the most common electrical vehicle (EV) chargers employ a two-stage design, i.e., a front-end AC/DC stage + an isolated DC/DC converter. In this paper, an isolated dual-active-bridge (DAB) based single-stage AC/DC converter was proposed, which has the power-factor-correction (PFC) and zero-voltage-switching (ZVS) functions over the full-load range. By reducing one power stage and eliminating the large DC link capacitor, a high efficiency and high power density are achieved. Such topology can be used as a modular building block to scale up to 50kW by serial connecting the input terminals and paralleling output terminals. A novel energy-balanced variable switching frequency control for such input-series-output-parallel (ISPO) modular designed is proposed. A single-phase d-q transformation is implemented to achieve zero steady-state error. Simulation analysis and experimental validation are presented.
Most of the present EV on-board chargers utilize a three-stage design, e.g., AC/DC rectifier, DC to high-frequency AC inverter, and AC to DC rectifier, which limits the wall-to-battery efficiency to ∼94%. To further increase the efficiency and power density, a matrix converter is an excellent candidate directly converting grid AC to high-frequency AC thereby saves one stage. However, its control complexity and the high cost of building the back-to-back switches are barriers its acceptance. Instead, this paper adopts the 650V E-mode GaN HEMTs to build a level-2 on-board charger using the indirect matrix topology. The input voltage is 80∼260VAC, the battery voltage is 200∼500VDC and the rated power is 7.2kW. Variable switching frequency is combined with phase-shift control to realize the zero-voltage switching. To further increase the system efficiency, four GaN HEMTs are paralleled to form one switching module with a novel gate-drive technology. The overall system efficiency is ∼98% and the power density is 3.3kW/L.
Full-bridge power-factor-correction (PFC) front-end + dual-active-bridge (DAB) AC/DC topology is widely used in industry, e.g., electrical vehicle on-board charger. Such two-stage topology limits the system efficiency, and the bulky DC link bus capacitor makes the system power density relatively low. Compared to the two-stage design, the single-stage design, unfolding bridge + DAB, eliminates the bulky DC link bus capacitor and operates the front-end with only 60Hz switching frequency, thereby has the potential to increase the system power density and efficiency. A novel variable-switching-frequency and hybrid single-dual-phase-shift (VSF-SDPS) control strategy is proposed and analyzed for the DAB based single-stage topology. The proposed VSF-SDPF control consists of two phase shifts to guarantee Zero-Voltage-Switching (ZVS) over the full range of the AC line voltage, and frequency modulation to achieve boost PFC. The conventional front-end PFC is simplified to an unfolding bridge by changing DAB control strategy to achieve PFC and ZVS at the same time. Besides, a special ZVS boundary is utilized to solve the grid current distortion problem when the switching frequency saturated, which is especially severe at light load condition. Simulation results and experimental validation are presented under 50Vrms AC line voltage and 200V DC battery voltage test condition.
Most of the present EV on-board chargers utilize a three-stage design, e.g., AC/DC rectifier, DC to high-frequency AC inverter, and AC to DC rectifier, which limits the wall-to-battery efficiency to ~94%. Instead of using the regular three-stage design, a matrix converter could directly convert grid AC to high-frequency AC thereby saves one stage and potentially increases the system efficiency, however, the control will be more complex and the high cost of building the back-to-back switches is inevitable. This paper adopts the 650V E-mode GaN HEMTs to build a level-2 on-board charger. The input voltage is 80~260VAC, the battery voltage is 200~500VDC and the rated power is 7.2kW with the bidirectional power-flow capability. Such design saves the bulky DC-bus capacitor. Variable switching frequency is combined with phase-shift control to realize the zero-voltage switching. An active filter is employed to choke the 120Hz output current ripple if needed. To further increase the system efficiency, four GaN HEMTs are paralleled to form one switching module. The overall system efficiency is >97% and the power density is 2.5kW/L with the active filter and 3.3kW/L without the active filter.
Conventional charging systems for electric and plug-in hybrid vehicles currently use cables to connect to the grid. This methodology creates several disadvantages, including tampering, risk, depreciation and non-value added user efforts. Loose or faulty cables may also create a safety issue. Wireless charging for electric vehicles delivers both a simple, reliable and safe charging process. The system enhances consumer adoption and promotes the integration of electric vehicles into the automotive market. Increased access to the grid enables a higher level of flexibility for storage management, increasing battery longevity. The power class of 3.7kW or less is an optimal choice for global standardization and implementation, due to the readily available power installations for potential customers throughout the world. One of the key features for wireless battery chargers are the inexpensive system costs, reduced content and light weight, easing vehicle integration. This paper demonstrates a wireless charging design with minimal component content. It includes a car pickup coil with 300 mm side length and low volume and mass 1.5 dm3 power interface electronics. After an overview of its hardware requirements, power transfer and efficiency benefits are presented, providing the anticipated horizontal and lateral deviations. An intense magnetic field is required to transfer the target power at low volumes between the transfer units. This field heats up any metal object over the transfer coil, similar to an induction oven. Consequently, the system should be powered down whenever a metal object is detected in this area. A Foreign Object Detection (FOD) design has been developed to continuously monitor the critical high field area. Device testing results are also provided.
Wide-Bandgap devices are believed to be promising candidates for the next-generation power electronics converter. However, the prohibitive cost and limited variety are still the main constrains before being widely used. One possible solution to mitigate these issues is hybrid switches, a combination of Si MOSFETs and WBG devices. To maximize merits of both GaN HEMT and Si MOSFET, this paper proposes a hybrid switch consisting of two GaN HEMTs and two Si MOSFETs. For such a design, a robust gate-drive loop is critical to secure the safe operation of all switches. As an instructive work, this paper presents a comprehensive analysis of the switching transient process and its impact on the gate-drive loop, practical tuning tips of gate-drive loop design are also given based on the simulation and experimental results. A 400V/ 80A full bridge prototype is developed to validate our design. Experimental result shows that, with enhanced gate-drive loop, we can continuously turn off 400V/80A@100kHz and 400V/40A@300kHz with only one GaN device paralleled to two commercial Si MOSFETs.
To filter the 120Hz output current ripple in our previously designed 7.2kW single-phase EV charger, this paper proposes to equip the charger with a buck-type active filter. 650V/60A enhancement mode GaN HEMTs provided by GaN Systems Inc are adopted to work at hard-switching mode. Experimental results indicated that four such switches could be paralleled to hard switch on/off ~240A, which is the key for the buck-type active filter. A model-based proportional-resonant controller is adopted to smooth the output current. Such control will enhance the dynamic response of the active filter, compared to the conventional PI controller. The experimental output current ripple and power loss analysis are given.
Wide-bandgap (WBG) devices are believed as the alternate of silicon switches for high-efficiency and high-power-density power electronics converters. While two major challenges of WBG devices remain as high cost (~5 times of Si) and less options (the maximum power rating for GaN is only 650V/60A), paralleling GaN with Si could be the potential solution to solve pains above. In this paper, two SMT GaN HEMTs are paralleled to a TO-247 Si MOSFET. A time delay is added between switch gate signals to make GaN endure the switching loss and Si conduct majority of the current, which maximizes advantages of both switches. The proposed design is found particularly useful for zero-voltage-switching applications. Critical dynamic behaviors such as the current overshoot to the GaN, current distribution during the dead time, and voltage spike during the turn-off caused by parasitics are comprehensively discussed. Its impact on the control performance and system loss is evaluated as well.
This paper designed the gate driver circuits and optimized the PCB layout in a 7.2kW battery charger using paralleled GaN HEMTs. 650V/60A enhancement mode GaN HEMTs provided by GaN Systems Inc are adopted. To optimize the switching performance of paralleled GaN HEMTs with low loss and high reliability, effects of parasitic inductance and capacitance are modeled and analyzed. Through cancelling the flux in the commutation loop, the power-loop parasitic inductance is reduced to only 0.7nH, which significantly decreases the electrical stress in the switch turn-off process. A diverse-parameter gate driver design has been proposed to achieve the reliable switching off. The Finite-Element-Analysis and Spice simulation show our current design could effectively suppress the voltage overshoot and gate-drive ringing on HEMTs. Experiments were carried out on both double pulse test platform and the 7.2kW charger to verify the proposed design strategy.
Global CO₂ reduction by 2021, according to some projections, will be comprised of multiple vehicle technologies with 7% represented by hybrid and electric vehicles (2% in 2014) [1]. Other low cost hybrid methods are necessary in order to achieve widespread CO₂ reduction. One such method is engine-off coasting and regenerative braking (or recuperation) using a conventional internal combustion engine (ICE). This paper will show that a 48V power system, compared to a 12V system with energy storage module for vehicle segments B, D and E during WLTP and NEDC, is much more efficient at reducing CO₂. Passive engine-off coasting using 12V energy storage shows a CO₂ benefit for practical real world driving, but, during NEDC, multiple sources of friction slow the vehicle down to the extent that the maximum benefit is not achieved. By adding active engine-off coasting at the 48V level the CO₂ emissions for NEDC are improved by decreasing the rate of deceleration with a 48V electric motor for propulsion. Also important, which will be explored in more detail, are the necessary power dimensions for the major components for different electrical load profiles.
Current significant challenges in the automotive industry for increasing fuel economy and reducing CO₂ emissions remain with traditional combustion engines. Moderately small increases in fuel efficiency lead to major reductions in CO₂ emissions, primarily due to large production volumes utilizing incremental fuel saving technologies. Enhancements of today’s vehicle powertrains, including micro-hybrids and mild-hybrids with stop-start systems, and coasting and energy recuperation have shown a positive cost benefit and shorter payback period. This is identified when the technology is compared to more complex and expensive HEVs (Hybrid Electric Vehicles) and BEVs (Battery Electric Vehicles). This paper describes the development of a baseline conventional vehicle model for estimating fuel savings and CO₂ reduction; it provides a benchmark for the development of fuel saving energy management technologies such as stop-start, coasting, and dual voltage architecture with regenerative braking and “on-demand” fuel senders. It will be shown that a stop-start system will provide a simulated 2.9% FE (Fuel Economy) benefit for the EPA unadjusted combined city/highway driving cycles. Also enhanced stop-start with aggressive coasting with engine-off (≺100 km/hr) provides an additional benefit of 7.1%. In addition, this paper describes a case study for the development of a HIL (Hardware-In-the-Loop) simulator which makes use of the conventional baseline model. The HIL system measures fuel savings of replacing a “100% driven” fuel system with an “on-demand” fuel delivery system. The case study will show a 40% CO₂ reduction over “100% driven” DC pump with a DC “on-demand” pump and an additional 22% CO₂ reduction for the BLDC “on-demand” pump for the EPA city/highway driving cycles using a Mini Cooper vehicle model.
Enhancements of today’s Micro-Hybrids based on stop-start systems with and without coasting and energy recuperation show a positive cost-benefit and a much shorter payback period compared to more complex and expensive Full-Hybrid concepts. However, improved Micro-Hybrid functionalities have a higher demand on the vehicle’s electrical power network, which cannot be covered with traditional topologies alone. To enable the advanced Micro-Hybrid features, additional energy storage elements like second lead acid batteries, double-layer capacitors or lithium-ion cell based storage systems will be integrated into the power network. This will stabilize the network and provide a reliable source of energy. To apply even further reaching measures like creeping (also called crawling), and high power recuperation, a dual voltage power network will be required. This can be achieved by adding a second voltage level to the traditional 12V power network. In order to connect power networks with different voltage levels, DC/DC converters are required. This paper will discuss the constraints of a cost-optimum topology for each power class of DC/DC converter from Micro- up to Mild-Hybrid applications and present solutions to overcome these limitations. Furthermore, the different topologies are compared in regards of their economic benefit using the total cost of ownership model.
The fierce competition and shifting consumer demands require automotive companies to be more efficient in all aspects of vehicle development and specifically in the area of embedded engine control system development. In order to reduce development cost, shorten time-to-market, and meet more stringent emission regulations without sacrificing quality, the increasingly complex control algorithms must be transportable and reusable. Within an efficient development process it is necessary that the algorithms can be seamlessly moved throughout different development stages and that they can be easily reused for different applications. In this paper, we propose a flexible engine control architecture that greatly boosts development efficiency. The efficiency improvements are achieved through four key features, namely the separation of target-dependent and target-independent algorithms, the partitioning of the control system at the system-level, the use of hardware I/O blocksets to facilitate system-level automatic coding, and the use of a data pooling concept to simplify signal propagation throughout the architecture layers. With these key features, the architecture supports seamless transitions from a concept to the final implementation stage and allows algorithm reuse and rapid algorithm substitution.
In GM R&D Powertrain/Engine Control Group, rapid prototyping controller (RPC) systems with Matlab/Simulink are used extensively to design, simulate and implement advanced engine control algorithms and models. However, those RPC systems use powerful microprocessors with large amounts of RAM contrary to engine control modules (ECM) in production vehicles. Therefore, a thorough analysis on the comparatively much more complicated algorithms and models cannot be performed during the research stage, since there are not enough tools to enable the smooth transition from Matlab/Simulink to the production type processor. The Real-Time Interface (RTI) Blockset for a production like microprocessor would close the transition gap between rapid prototyping controller systems and production type microprocessors by leveraging the power and popularity of Matlab/Simulink in control engineering world and automatic code generation tools. Therefore, RTI Blockset with different customized functionalities for a production like microcontroller is developed in Matlab/Simulink in order to facilitate on-target rapid prototyping. Moreover, an engine control algorithm, which was developed before on a production like microcontroller via handwritten C-code is implemented on the same hardware using Matlab/Simulink Embedded Coder and the aforementioned RTI Blockset.
Combustion feedback using cylinder pressure sensors, ion current sensors or alternative sensing techniques is actively under investigation by the automotive industry to meet future legislative emissions requirements. One of the drawbacks of many rapid prototyping engine management systems is their available analog interfaces, often limited to 10-12 bits with limited bandwidth, sampling rate and very simple anti-aliasing filters. Processing cylinder pressure or other combustion feedback sensors requires higher precision, wider bandwidths and more processing power than is typically available. For these reasons, Ricardo in collaboration with GM Research has developed a custom, high precision analog input subsystem for the rCube rapid prototyping control system that is specifically targeted at development of combustion feedback control systems. Up to 32 channels of high precision, wide bandwidth analog signals are pre-processed by a Xilinx Spartan III Field Programmable Gate Array before being transferred via dual port memory to a high performance PowerPC for further processing. The design enables that hardware and software requirements for combustion feedback systems to be investigated without effective limitations of the analog system and processing power. Requirement specifications for production systems can therefore be developed in a systematic manner. Rapid prototyping of control algorithms can be performed in either ASCET-SD or Simulink. This paper describes the hardware design and implementation of the rapid prototyping system for investigation of combustion feedback control technology for both gasoline and diesel systems.
A rapid prototyping controller (RPC) based, full-authority, diesel control system is developed, implemented, tested and validated on FTP cycle. As rapid prototyping controller, dSPACE Autobox is coupled with a fast processor based slave for lower level I/O control and a collection of in-house designed interface cards for signal conditioning. The base software set implemented mimics the current production code for a production diesel engine. This is done to facilitate realistic and accurate comparison of production algorithms with new control algorithms to be added on future products. The engine is equipped with all the state-of-the art subsystems found in a modern diesel engine (common rail fuel injection, EGR, Turbocharger etc.).
A frequency-domain approach to balancing of air-fuel ratio (A/F) in a multi-cylinder engine is described. The technique utilizes information from a single Wide-Range Air-Fuel ratio (WRAF) or a single switching (production) O₂ sensor installed in the exhaust manifold of an internal combustion engine to eliminate the imbalances. At the core of the proposed approach is the development of a simple novel method for the characterization of A/F imbalances among the cylinders. The proposed approach provides a direct objective metric for the characterization of the degree of A/F imbalances for diagnostic purposes as well as a methodology for the control of A/F imbalances among various cylinders. The fundamental computational requirement is based on the calculation of a Discrete Fourier Transform (DFT) of the A/F signal as measured by a WRAF or a switching O₂ sensor. For real-time applications, the approach is iterative in nature and the intended goal of A/F balancing is achieved quite accurately and fast in around 1-2 seconds. Experimental results for 3-cylinder and 6-cylinder applications are provided and the advantages and limitations of the technique including the effect of blind spots pointed out. Application of the technique will result in lower tailpipe emissions on FTP and US06 driving cycles as well as tangible (but minor) fuel economy gains. Clearly, in extreme cases of cylinder A/F maldistributions, the impact on the maintenance of good driveability is of paramount importance and achieved automatically using the technique on-board the vehicle in its real-world operation.
Phasellus non ante ac dui sagittis volutpat. Curabitur a quam nisl. Nam est elit, congue et quam id, laoreet consequat erat. Aenean porta placerat efficitur. Vestibulum et dictum massa, ac finibus turpis.
