新手E5061B用户需要注意哪些事项？

> {quote：title = Antonio Guerrieri写道：} {quote}>特别是，我不明白的是Avg，IF BW Auto，IFBW Auto限制的含义？
>我看到这些参数在某种程度上改变了扫描时间，在某些情况下也改变了频率跨度，很可能改变了实验所需的点数。
>>有人可以帮我吗？
>有人可以建议一些阅读吗？
>>很多非常感谢提前>> Antonio我不知道E5061B，所以以下是基于我使用其他VNA，频谱分析仪和锁相放大器的经验。
简单来说*降低IF带宽*增加平均值可以降低测量噪声。
如果使用较小的IF带宽，则会收到较少的噪声。
噪声功率将与IF带宽成比例。
在降低的IF带宽下，您可以获得更好的动态范围。
但是，滤波器将具有建立时间。
这通常取决于很多事情（响应，顺序等），但与带宽成反比。
因此，IIF带宽越窄，滤波器所需的时间就越长，因此扫描必须越慢。
随着您查看的数据点数量的增加，必须收集更多数据点，过滤器允许更多次结算，因此扫描需要更长时间。
在我的VNA中至少（8720D）并且我期望在很多VNA中，频率范围被分成频段。
在频带之间切换需要一些时间，因此对于相同数量的数据点，相同的IF带宽，扫描速度不一定相同。
为了无耻地复制Dr_Joel前几天写的内容，当平均N个数据点时，使用的方法是：New_Data_Displayed =（New_Data / N）+ Old_displayed_data *（N-1）/ N执行无限的impule响应过滤器。
显而易见的是，随着平均更多数据，噪声会减少。
在从VNA读取读数之前，您还需要让仪器稳定一段时间。
对于某些测量，例如在源和接收器之间存在大量衰减，您需要降低噪声并接受更长的测量时间。
对于其他情况，这不是必需的。
这些平均，IF带宽和噪声影响问题对频谱分析仪等其他仪器来说相当普遍。
通常，扫描速度根据点数和IF带宽自动设置。
您通常可以覆盖默认值，但通常不需要很多。
我知道需要做的一个例子是在一些旧的VNA上使用长电缆。
应用笔记+“10个用于实现更好的网络分析仪测量的提示”+，应用笔记1291-1B中对此进行了描述。
另请参阅+“理解和改进网络分析仪动态范围”Dave



     以下为原文

  > {quote:title=Antonio Guerrieri wrote:}{quote}

> Particularly,what I don't understand is the meaning of Avg, IF BW Auto, IFBW Auto limit?
> I see that these parameters in some way change the scan time as well as in some cases the frequency span and much probably the number of points which are acquisite with the experiment.
> 
> Can someone help me about?
> Can someone suggest some reading about?
> 
> many many thanks in advance
> 
> Antonio

I don't know the E5061B, so the following is based on my experience using other VNAs, spectrum analyzers and lock-in amplifiers. 

In simple terms

* Reducing the IF bandwidth
* Increasing the average 

both reduce the noise on the measurement. 

If you use a smaller IF bandwidth, less noise is receiveed. Noise power will be proportional to IF bandwidth. You get a better dynamic range at reduced IF bandwidth. 

However, the filters will have a settling time. That in general depends on many things (response, order etc), but is inversely proportiaal to the bandwidth. So the narrower the IIF bandwidth, the longer the filters take to settle, so the slower the sweep must be. 

As the number of data points you look at increaseed, so more data points must be collected, the filter allowed to settle more times, so the sweep takes longer. 

In my VNA at least (8720D) and I expect in a lot of VNAs, the frequency range is split into bands. It takes some time to switch between bands, so for the same number of data points, the same IF bandwidth, the sweep speed will not necessarily be the same. 

To unashamedly copy what Dr_Joel wrote the other day, when averaging N data points, the method used is:

New_Data_Displayed=(New_Data/N)+Old_displayed_data*(N-1)/N

That performs an infinite impule response filter. It should be obvious that as you average more data the noise gets reduced. You also need to let the instrument settle for longer before taking readings off of the VNA. 

For some measurements, where for example there is a lot of attenuation between the source and the receiver, you need to reduce noise and accept a longer measurement time. For other cases, this is not necesssary. 

These issues of averaging, IF bandwidth and noise effects are fairly common to other instruments like the spectrum analyzer. 

In general, the sweep speed is set automatically depending on the number of points and the IF bandwidth. You can generally override the defaults, but there is often not a lot of need to. One example I do know of where it needs to be done is in using long cables on some older VNAs. That is described in the application note +"10 Hints for Making Better Network Analyzer Measurements"+, Application Note 1291-1B. Also take a look at +"Understanding and Improving Network Analyzer Dynamic Range"



Dave

非常感谢Dave现在我开始理解了一些......可能你引用的文献可以帮助我更多......我正在做阻抗测量随着时间的推移，因为我对石英晶体属性如何变化感兴趣
时间，所以我需要控制或至少知道每次测量的扫描时间.....希望你的回复可以帮助再次感谢安东尼奥



     以下为原文

  Thank so much Dave

now i began to understand some ......

probably the literature you cited can help me much more ....

I'm doing impedance measurements with time since i'm interested how the quartz crystal properties changes with time so that i need to control or at least to know the scan time for each measurements .....

hope your reply can help

thanks so much again 

Antonio

> {quote：title = Antonio Guerrieri写道：} {quote}>非常感谢Dave那没问题安东尼>我正在做阻抗测量随着时间的推移因为我对石英晶体属性如何随时间变化感兴趣所以我需要
控制或至少知道每次测量的扫描时间......石英非常稳定。
我不确定你期望在VNA的扫描时间内看到什么变化。
但至少在我使用的任何VNA上，都可以控制扫描时间。
您可能需要考虑订阅+ time-nuts +邮件列表。
http://leapsecond.com/time-nuts.htm我怀疑你会得到很多关于水晶的有用建议。
+时间坚果+的几个订户有多个铯振荡器。
至少有一个（Tom Van Baak）有一个氢脉冲时钟 - 比铯钟更稳定（而且价格昂贵！）。
汤姆已经将铯钟带到山上进行实验，以验证爱因斯坦的相对论。
有关石英振荡器的+ time-nuts +列表中有很多* expertese。
汤姆网站上的一些更令人惊叹的东西是一个非常精确的手表：http：//leapsecond.com/pages/atomic-bill/（安捷伦员工应该特别赞赏）BNC国际象棋http://leapsecond.com/
页/象棋/戴夫



     以下为原文

  > {quote:title=Antonio Guerrieri wrote:}{quote}
> Thank so much Dave

That's no problem Antonio

> I'm doing impedance measurements with time since i'm interested how the quartz crystal properties changes with time so that i need to control or at least to know the scan time for each measurements .....

Quartz is pretty damm stable. I'm not sure what changes you expect to see during the sweep time of a VNA. But at least on any VNA I have used, it is possible to control the sweep time. 
 
You might want to consider subscribing to the +time-nuts+ mailing list. 

http://leapsecond.com/time-nuts.htm

I suspect you would get a lot of useful advice about crystals. 

Several of the subscribers to +time-nuts+ have multiple cesium oscillators. At least one (Tom  Van Baak) has a hydrogen maser clock - which is more stable (and expensive!) than a cesium clock.  Tom has taken cesium clocks up mountains for experiments to verify Einstein's theory of relativity. 

There is a *lot* of expertese on the +time-nuts+ list about quartz oscillators. 

Some of the more amuzing things on Tom's site are a really accurate wrist watch:

http://leapsecond.com/pages/atomic-bill/

(which should be especially appreciated by Agilent staff)

BNC chess

http://leapsecond.com/pages/chess/


Dave