Quietening ultra-low-loss SiC & GaN waveforms

电力电子减少了我们的碳足迹，每年为英国经济贡献近500亿英镑，并为400多家英国公司提供82，000个技术工作岗位。电力电子转换器调节大多数电气设备中的功率流，在电动汽车等中，它们通过每秒数千次的开关电流来实现，并且接通时间与关断时间的比率决定了功率流。这些转换器的效率、尺寸和重量取决于产生的废热量。例如，笔记本电脑电源适配器的尺寸多年来已经缩小，这是由于它们的效率提高。在电动汽车中，废热导致功率转换器通常比它们所提供的电机更大。这种热量主要是在晶体管切换时产生的。电力电子行业即将发生重大变化，因为最近出现了由碳化硅（SiC）或氮化镓（GaN）制成的超快转换晶体管。它们的开关转换非常短（低于10纳秒），原则上，效率可以被推到以前从未达到的水平。这可能会导致十倍的小型化，导致转换器比驱动的电机小得多，或者信用卡大小的笔记本电脑电源适配器。然而，快速开关伴随着极端电磁噪声的缺点，行业正在努力采用这些新技术。我们的项目将为采用这些新技术的关键不确定性提供答案，即如何快速，安全和安静地驱动SiC和GaN功率器件。电磁噪声（EMI）在示波器上被视为尖锐的角落，快速振荡和过冲尖峰，在开关过渡期间。在这个项目中，我们正在开发解决方案，以实现清洁开关，没有这些不良特性，让EMI安静下来这些特征通过将特殊形状的信号馈送到晶体管的栅极来抵消。对于任何已知的信号发生器和闭环控制方法或无源开关辅助（滤波）电路来说，开关转换太快，无法提供所需的门信号整形。因此，我们采用了另一种方法，我们最近开发了一种芯片，它可以每100皮秒（即光传播3厘米所需的时间）调整一次电流输出。它是唯一已知的驱动芯片，可以与栅极信号足够频繁地相互作用，以形成这些短的10纳秒以下的开关转换。我们将创建此驱动器的改进版本，以驱动氮化镓和碳化硅晶体管栅极，其信号旨在软化开关并消除不必要的高频效应。当转换器温度改变时，以及当要求改变其输出功率时，信号需要自动改变。此外，每种类型的电路需要略有不同的信号。因此，将开发自动适应，以简化工业界对这项技术的使用。一个有趣的挑战是安全生成优化的栅极信号，因为错误的信号可能导致电源转换器故障。另一个挑战是输入门的能量的再生，以便它可以用于下一个开关事件。该项目开发微电子（高速，EMI安静门驱动器）和电力电子（转换器和控制系统）。来自8家合作伙伴公司的行业顾问将指导三年的发展。在第四年，该研究将缩小规模，并在英国的工业中进行试验，以转移专有技术，测试研究，并为基础研究提供新的途径。这项研究将有助于保持新兴的高效率电力电子和现代超低功耗微电子之间的兼容性，这对电磁噪声越来越敏感，并简化和加快SiC和GaN的工业采用。

Power electronics reduces our carbon footprint and contributes nearly £50bn per year to the UK economy and supports 82,000 skilled jobs in over 400 UK-based companies. Power electronic converters regulate the flow of power in most electrical devices, in electric vehicles etc. They do so by switching currents on and off, 10s of thousands of times per second, and the ratio of on-time to off-time determines the power flow. The efficiency, size, and weight of these converters are determined by the amount of waste heat generated. For example, the size of laptop power adapters has shrunk over the years, due to their increase in efficiency. In an electric car, waste heat causes power converters to be typically larger than the motors they are feeding. This heat is mostly produced in the instances when the transistors are switching. The power electronics industry is about to undergo significant change, as ultra-fast-transition transistors made from silicon carbide (SiC) or gallium nitride (GaN) have recently emerged. Their switching transitions are so short (below 10 nanoseconds) that, in principle, efficiency can be pushed to levels never achieved before. This could lead to a ten-fold miniaturisation, leading to converters that are much smaller than the motor being driven, or credit-card-sized laptop power adapters.The fast switching, however, comes with the downside of extreme electromagnetic noise, and industry is struggling to adopt these new technologies. Our project will provide answers to key uncertainties for adoption of these new technologies, namely how to drive the SiC and GaN power devices quickly, safely and quietly.The electromagnetic noise (EMI) is seen on an oscilloscope as sharp corners, rapid oscillations, and overshoot spikes, during the switching transitions.In this project, we are developing solutions to achieve clean switching, without these undesirable features, to quieten the EMI. These features are countered by feeding specially-shaped signals into the transistors' gates. The switching transition is too fast for any known signal generators and closed-loop control methods, or passive switching-aid (filtering) circuits to provide the required shaping of gate signals. Therefore, an alternative approach is adopted.We recently developed a chip that can adjust its current output every 100 picoseconds, i.e. the time it takes light to travel 3 cm. It is the only known driver chip that can interact frequently enough with a gate signal to shape these short sub-10 nanosecond switching transitions. We will create improved versions of this driver to drive gallium nitride and silicon carbide transistor gates with signals that are designed to soften the switching and cancel out unwanted high-frequency effects. The signals need to be changed automatically as the converter temperature changes, and when changes to its output power are requested. Also, each type of circuit requires slightly different signals. Therefore, automatic adaptation will be developed to simplify the use of this technology by industry. An interesting challenge is the safe generation of optimised gate signals, as the wrong signal can cause a power converter to fail. Another challenge is the regeneration of energy put into the gate, so that it can be used for the next switching event.The project develops microelectronics (high-speed, EMI-quietening gate drivers) and power electronics (converters and control systems). Industry advisors from 8 partner companies will steer the development for three years. In Year 4, the research is scaled down, and trials in UK-based industry set up to transfer knowhow, test the research, and provide new avenues for fundamental research.This research will help maintain the compatibility between emerging high-efficiency power electronics and modern ultra-low-power microelectronics that is increasingly susceptible to electromagnetic noise, and simplify and expedite industry adoption of SiC & GaN.