当前位置:网站首页>电源噪声分析
电源噪声分析
2022-06-24 18:32:00 【Risehuxyc】

电源噪声分析
一提到电源噪声,相信就会引起很多电子工程师的共鸣。我们平时所说的电源噪声到底是什么呢?它等同于电源纹波吗?事实上,电源噪声不同于电源纹波,它是出现在输出端子间的纹波以外的一种高频成分。而纹波是出现输出端子间的一种与输入频率、开关频率同步的成分,是叠加在稳定直流信号上的交流干扰信号。

电源噪声波形
在电源噪声的分析过程中,比较经典的方法是使用示波器观察电源噪声波形并测量其幅值,据此判断电源噪声的来源。但是随着数字器件的电压逐步降低、电流逐步升高,电源设计难度增大,在观察时域波形无法定位故障时,可以通过 FFT(快速傅立叶变换)方法进行时频转换,将时域电源噪声波形转换到频域进行分析。电路调试时,从时域和频域两个角度分别来查看信号特征,可以有效地加速调试进程。
示波器的频域分析功能是通过傅立叶变换实现的,傅立叶变换的实质是任何时域的序列都可以表示为不同频率的正弦波信号的无限叠加。我们分析这些正弦波的频率、幅值和相位信息,就是将时域信号切换到频域的分析方法。数字示波器采样到的序列是离散序列,所以我们在分析中最常用的是快速傅立叶变换(FFT)。 FFT算法是对离散傅立叶变换(DFT)算法优化而来,运算量减少了几个数量级,并且需要运算的点数越多,运算量节约越大。
示波器捕获的噪声波形进行FFT变换,有几个关键点需要注意:
1、根据耐奎斯特抽样定律,变换之后的频谱展宽(Span)对应与原始信号的采样率的1/2,如果原始信号的采样率为1GS/s,则FFT之后的频谱展宽最多是500MHz;
2、变换之后的频率分辨率(RBW Resolution Bandwidth)对应于采样时间的倒数,如果采样时间为10mS,则对应的频率分辨率为100Hz;
3、频谱泄漏,即信号频谱中各谱线之间相互干扰,能量较低的谱线容易被临近的高能量谱线的泄漏所淹没。避免频谱泄漏可以尽量采集速率与信号频率同步,延长采集信号时间及使用适当的窗函数。
电源噪声测量时要求采集的信号时间可以足够长,可以认为覆盖到了整个有效信号的时间跨度。
边栏推荐
- Selection (032) - what is the output of the following code?
- 云服务器类型怎么选,要考虑什么?
- What is business intelligence (BI)?
- variable
- Wechat applet development - Implementation of rotation chart
- Six configuration management tools that administrators must know
- JDBC writes Chinese garbled code to the database
- About whether arm's large and small end mode is related to CPU or compiler
- Several key points for enterprises to pay attention to digital transformation
- Sudoku (easy to understand)
猜你喜欢
Paper sharing | self supervised learning paper jointly released by Yann Lecun and read by engineers

Architecture decryption from distributed to microservice: several common microservice architecture schemes

微服务系统设计——子服务项目构建

JS event details

SAP license: ERP for supply chain management and Implementation

How do yaml files and zmail collide with the spark of the framework, and how can code and data be separated gracefully?

Flutter dart regular regexp special characters $, () (IV)

13 ways to reduce the cost of cloud computing

JS deep understanding of functions

How can an enterprise successfully complete cloud migration?
随机推荐
如何在 R 中执行幂回归
(Video + graphics) introduction to machine learning series - Chapter 11 support vector machines
Graph traversal (BFS and DFS) C language pure handwriting
微服务系统设计——数据模型与系统架构设计
variable
Get max value of a bit column - get max value of a bit column
Bigdecimalavoiddoubleconstructorrule: do not directly use the double variable as a parameter to construct BigDecimal
Millions of dollars worth of NFT were stolen in the attack, and Google issued an emergency warning to 3.2 billion users worldwide | February 21 global network security hotspot
JS event details
Why are more and more people studying for doctors? Isn't it more and more difficult to graduate a doctor?
[golang] leetcode intermediate - jumping game & different paths
R language Quantitative Ecology redundancy analysis RDA analysis plant diversity species data visualization
13 skills necessary for a competent QA Manager
[NLP] 3 papers on how Stanford team builds a better chat AI
Why should state-owned enterprises accelerate the digital transformation
Keep two decimal places
JS clothing photo album case
Digital trend analysis of B2B e-commerce market mode and trading capacity in electronic components industry
Restcloud ETL extracting dynamic library table data
Number of occurrences of numbers in the array (medium difficulty)