Low-noise Design for Power Delivery Network in 2.5D Integrated Microsystem

作者
Yanling Wang,Qiao Yang,Liang Zhao,Baoxia Li
出处
期刊: 卷期号:: 1-4 被引量:1
标识
DOI:10.1109/icept59018.2023.10492209
摘要

Power delivery network (PDN) plays an important role in high-density integrated microsystem package to supply clean and stable power. High speed synchronous switch, shared silicon substrate and small integrated space always cause large transient changes in current and EMI, which make the design of PDN more difficult. This paper focuses on the core power supply design of the CPU in 2.5D integrated microsystem to realize low noise PDN, whose impedance must be less than 0.1Ω at frequencies below 5GHz. The researched PDN is composed of TSV-interposer PDN and HTCC-substrate PDN. An effective methodology called chip-interposer-substrate co-design and decoupling capacitors (de-cap) co-optimization is presented. For decoupling capacitors, deep trench capacitors (DTC) are used in TSV interposer and surface mount capacitors (SMC) are used on HTCC substrate. First, the decoupling effect of DTCs in TSV interposer and SMCs on HTCC substrate are compared for the whole microsystem PDN in term of frequency domain. Meanwhile, three kinds of capacitance and distributions of DTCs in TSV interposer are also discussed. It carries out that DTCs in TSV interposer can obtain flat impedance in wide frequency band, especially frequencies upper than 100MHz. From 100MHz to 5GHz, the impedance of PDN with multiple small DTCs is maximally reduced by 4Ω than SMCs. Furthermore, the low voltage noise generated by DTCs is verified in term of time domain. Circuit is built using PDN’s S parameter and chip power model (CPM) and transient voltage analysis is done. For universality, CPM models of three CPU chips are discussed. All the three results show that DTCs can significantly improve the PDN performance. PDN with DTCs compared to SMCs, the maximum voltage noise is reduced by 385.6mV.

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
reeal发布了新的文献求助10
1秒前
1秒前
咚咚咚发布了新的文献求助10
1秒前
1秒前
Jello发布了新的文献求助10
1秒前
1秒前
vampv的应助被听雨采纳,获得10
2秒前
2秒前
2秒前
vistagy完成签到 ,获得积分10
3秒前
3秒前
jiangshan完成签到,获得积分10
3秒前
英俊的铭的应助被天一采纳,获得10
3秒前
碘伏发布了新的文献求助10
3秒前
充电宝的应助被徐小新采纳,获得10
4秒前
4秒前
5秒前
呆萌绿海完成签到 ,获得积分10
6秒前
Whalefall完成签到,获得积分10
6秒前
凉瑾完成签到 ,获得积分10
7秒前
7秒前
Zero完成签到,获得积分10
7秒前
Owen的应助被式微采纳,获得20
7秒前
张小墨发布了新的文献求助10
7秒前
7秒前
Zn9032完成签到,获得积分10
7秒前
灵药完成签到 ,获得积分10
8秒前
30完成签到 ,获得积分10
8秒前
8秒前
耶斯发布了新的文献求助10
9秒前
在水一方的应助被果冻采纳,获得10
9秒前
9秒前
赫连烙完成签到,获得积分0
9秒前
尊敬的寄松的应助被实验室采纳,获得10
9秒前
9秒前
10秒前
WIND-CUTTER完成签到,获得积分10
10秒前
10秒前
cdercder的应助被科研通管家采纳,获得10
10秒前
Akim的应助被科研通管家采纳,获得10
10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
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1500
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology)求助 1000
Encyclopedia of Modern Artificial Intelligence 1000
John McDowells Bildungstheorie 800
Analyses de reconnaissances de dettes passées devant les échevins d'Ypres (1249-1291) 600
Synthesis, Application and Future Perspectives of Smart Nano-materials - Part 1 400
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8153389
求助须知:如何正确求助?哪些是违规求助? 9550014
关于积分的说明 21696271
捐赠科研通 7741851
什么是DOI,文献DOI怎么找? 3425440
关于科研通互助平台的介绍 2550608
邀请新用户注册赠送积分活动 2469123