在线咨询
eetop公众号 创芯大讲堂 创芯人才网
切换到宽版

EETOP 创芯网论坛 (原名:电子顶级开发网)

手机号码,快捷登录

手机号码,快捷登录

找回密码

  登录   注册  

快捷导航
搜帖子
查看: 15522|回复: 20

[求助] PLL的杂散和相噪怎么理解

[复制链接]
发表于 2015-9-28 14:21:25 | 显示全部楼层 |阅读模式

马上注册,结交更多好友,享用更多功能,让你轻松玩转社区。

您需要 登录 才可以下载或查看,没有账号?注册

x
如图1所示,参考杂散为参考频率处的功率和载波处的功率差,单位为dBc;图2中,相位噪声曲线上的一些脉冲凸起为杂散,单位为dBc/Hz; 单从单位来看,这两个数值的差别为一个是功率差,一个表示单位单宽内的功率差值,那么这两者是怎样转换的呢,因为根据杂散的计算公式得到的单位为dBc;例如,计算得到800KHz 处的杂散要小于-73dBc,那么800KHz 表示的应该是频偏800KHz,即第二张图得到的杂散值吧,但是图2 的曲线不能通过仿真得到,根据计算的结果,怎么转换为图1中的值呢。另外,根据相噪的定义可以从图1中计算得到相位噪声吗??
求解答,谢谢!

图2

图2

杂散计算公式

杂散计算公式
 楼主| 发表于 2015-9-28 14:22:13 | 显示全部楼层
图1.jpg 上面是图2  这楼是图1
发表于 2015-9-28 16:08:51 | 显示全部楼层
图2如果你把span做到1HZ一个点,应该和相噪的图就是一个东西了,我的个人理解。不知道对不对,但我一直这么用的的。
 楼主| 发表于 2015-9-28 16:34:00 | 显示全部楼层
回复 3# bright_pan


   你说的是图1吧,那不是要去很多的点,你这样做和用pnoise仿的结果是一致的吗
发表于 2017-7-30 11:20:34 | 显示全部楼层
兄弟,话不多说,杂散抑制是相对值
发表于 2018-9-2 22:02:50 | 显示全部楼层
同问,已经用Cadence仿真得到PLL的输出频谱的杂散,那根据该图,应该如何计算频偏1MHz处的相位噪声呢?求教了!
发表于 2018-9-3 01:48:37 | 显示全部楼层
Spur tone is a deterministic tone. The unit of the spur is dBc, which is the ratio of the spur tone power to carrier (signal) power.  Phase noise is noise to signal ratio. The noise power is calculated in 1Hz bandwidth, consequently, you have the phase noise unit dBc/Hz. You can also calculate spur based on phase noise. This is related to the resolution bandwidth you are using. If the resolution bandwidth is 1Hz, then tone level in the phase noise plot is the spur.
发表于 2018-9-3 08:17:17 | 显示全部楼层
回复 7# gongjiang


   Thank you for your replying. Is the resolution bandwidth FFT used?If the resolution bandwidth is not 1Hz, then how to calculate the phase noise using the spur spectrum?
发表于 2018-9-3 21:14:35 | 显示全部楼层
回复 8# 电子新手是也
   Yes. The phase noise plot comes from the psd- fft of the auto-correlation function of the oscillation waveform (traditional definition, or the fft of the oscillation phase(IEEE definition) ). When you do the fft in MATLAB, you have the resolution frequency. Your phase noise plot is spread from 0 to fout/2 with the finite number of points (nr_points). The resolution frequency (fres) is fout/2/_nr_points. If fres is larger than 1, which means the spur value on phase noise plot is underestimated since the spur energy is the same but spread by fres. Consequently,  you can calculate the spur from PN plot by adding 10*log10(fres). But this is typically not the real case (I observed when using R&S) when you calculated from a phase noise analyzer since a phase noise analyzer has different fres at different frequency offset.
Regarding calculating phase noise from the spectrum, you can also do something similar. The spectrum analyzer is used to calculate the spectrum of the oscillation waveform instead of psd. You can calculate the noise to signal ratio from a spectrum at a certain offset frequency on the spectrum analyzer, then subtracting 10*log10(fres).  Basically, you calculated NSR at 1Hz bandwidth...
发表于 2018-9-6 22:56:14 | 显示全部楼层
回复 9# gongjiang


   Thank you very much~
您需要登录后才可以回帖 登录 | 注册

本版积分规则

关闭

站长推荐 上一条 /2 下一条

小黑屋| 关于我们| 联系我们| 在线咨询| 隐私声明| EETOP 创芯网
( 京ICP备:10050787号 京公网安备:11010502037710 )

GMT+8, 2024-4-27 12:11 , Processed in 0.034060 second(s), 9 queries , Gzip On, Redis On.

eetop公众号 创芯大讲堂 创芯人才网
快速回复 返回顶部 返回列表