A 48-V-to-1-V Switching Bus Converter for Ultra-High-Current Applications

电容器 占空比 拓扑(电路) 电子线路 电感器 功率(物理) 电压 计算机科学 开关电容器 电气工程 电子元件 电子工程 工程类 物理 量子力学
作者
Yicheng Zhu,Ting Ge,Nathan M. Ellis,Jiarui Zou,Robert C. N. Pilawa-Podgurski
标识
DOI:10.1109/compel52896.2023.10221149
摘要

This paper presents an ultra-high-current switching bus converter with direct 48-V-to-1-V power conversion for next-generation ultra-high-power digital loads (e.g., CPUs, GPUs, ASICs, etc.). In the proposed topology, two 2-to-1 switched-capacitor (SC) front-ends are merged with four 10-branch series-capacitor buck (SCB) modules through two switching buses. Compared to the DC-bus-based architecture, the switching-bus-based architecture does not require DC bus capacitors, reduces the number of switches, and ensures complete soft-charging operation. Through two-phase operation within each SCB module, the switching bus converter extends the maximum duty ratio and achieves a very large SC stage conversion ratio of 20-to-1. Compared to existing 48-V-to-1-V hybrid SC demonstrations, the proposed topology has the lowest normalized switch stress and the smallest normalized passive component volume, showing great potential for both higher efficiency and higher power density than prior solutions. A 48-V-to-1-V hardware prototype was designed and built with custom four-phase coupled inductors and gate drive daughterboards. Hybrid gate drive circuitry comprising gate-driven charge pump circuits and cascaded bootstrap circuits was customized for the high-side switches in the SCB modules to overcome the challenge of accumulative voltage drops in the conventional cascaded bootstrap circuit. The hardware prototype was tested up to 1200-A output current and achieved 92.4% peak system efficiency, 87.5% full-load efficiency (including gate drive loss), and 607 W/in 3 power density (by box volume).

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
new完成签到,获得积分10
1秒前
万能图书馆应助燕尔蓝采纳,获得10
2秒前
今后应助zzzzzz采纳,获得10
2秒前
蟹鱼橙子发布了新的文献求助10
2秒前
学术小牛发布了新的文献求助10
3秒前
3秒前
4秒前
Herowho完成签到,获得积分10
4秒前
李倩发布了新的文献求助30
4秒前
5秒前
5秒前
传奇3应助new采纳,获得10
8秒前
学术小牛发布了新的文献求助10
8秒前
9秒前
9秒前
Herowho发布了新的文献求助10
10秒前
10秒前
10秒前
GPTea发布了新的文献求助10
10秒前
11秒前
王十二发布了新的文献求助10
11秒前
香蕉觅云应助清爽的晓啸采纳,获得10
11秒前
积极乐观向上永不放弃的小孩完成签到,获得积分10
12秒前
彭于晏应助孤独采纳,获得10
12秒前
12秒前
香蕉觅云应助儒雅小蜜蜂采纳,获得10
12秒前
无极微光应助Mika采纳,获得20
13秒前
我是老大应助cindy采纳,获得10
13秒前
13秒前
luxiaoxi发布了新的文献求助10
13秒前
二氧化硒发布了新的文献求助10
14秒前
jiahui发布了新的文献求助10
14秒前
15秒前
难过龙猫发布了新的文献求助10
15秒前
衫青发布了新的文献求助10
16秒前
万能图书馆应助yangjun采纳,获得10
16秒前
16秒前
16秒前
儒雅芙蓉发布了新的文献求助10
16秒前
在水一方应助初君采纳,获得10
16秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Kinesiophobia : a new view of chronic pain behavior 2000
Research for Social Workers 1000
Mastering New Drug Applications: A Step-by-Step Guide (Mastering the FDA Approval Process Book 1) 800
The Social Psychology of Citizenship 600
Signals, Systems, and Signal Processing 510
Discrete-Time Signals and Systems 510
热门求助领域 (近24小时)
化学 材料科学 生物 医学 工程类 计算机科学 有机化学 物理 生物化学 纳米技术 复合材料 内科学 化学工程 人工智能 催化作用 遗传学 数学 基因 量子力学 物理化学
热门帖子
关注 科研通微信公众号,转发送积分 5912187
求助须知:如何正确求助?哪些是违规求助? 6831436
关于积分的说明 15785215
捐赠科研通 5037204
什么是DOI,文献DOI怎么找? 2711599
邀请新用户注册赠送积分活动 1661950
关于科研通互助平台的介绍 1603905