E5071C外部触发模式如何工作?

在扫描时，每个触发器提供一次完整扫描。
点上触发每个触发器一个数据采集点。
在具有N个点的2端口校准中，扫描需要2 * N个触发器。
假设您的脉冲比全扫描短得多，因此您必须测量每个脉冲的单个点。
实际上，您必须确保脉冲宽度比每个点的ENA数据采集时间长：为了简单起见，您可以使用IFBW> 1.1 /（脉冲宽度）。
您可能会发现，在某些情况下，您需要更宽的IF BW才能获得良好的测量结果。
然后您可以使用点模式平均或扫描模式平均来降低噪声（因为IF BW宽，噪声会更高）。



     以下为原文

  on sweep gives you one full sweep per trigger.  On point triggers one data acquisition point per trigger.  In a 2 port cal, with N points, you need 2*N triggers for the sweep.

Presumeably your pulse is much shorter than a full sweep, so you must measure a single point per pulse.  In fact, you must make sure that the pulse width is longer than the data acquisiton time of the ENA on a per point basis: to keep things simple you can use  IFBW>1.1/(pulse width).  You might find that for some cases, you need an even wider IF BW to get good measurements.  And then you can use point-mode averaging or sweep mode average to reduce the noise (since the IF BW will be wide, the noise will be higher).

谢谢Joel博士的回复。
我对NWAs很新，所以它听起来可能是一个愚蠢的问题......但是如果脉冲比全扫描短得多会发生什么？*例如扫描时间是250microsec，脉冲宽度是100微秒（脉冲周期1000microsec）
）。
在这种情况下，NWA会发生什么？
什么使数据捕获无效？
我想得到Joel博士或其他阅读这篇文章的人的帮助。
是否有任何特定的书可以给我关于NWAs的基本想法？
一种速成课程方法最适合我。
（我最近在图书馆找到了Joel博士的一本书）。
任何你作为参考/ *用来获得/进一步了解你的知识*都将受到高度赞赏。编辑：APL于2015年1月25日下午7:27



     以下为原文

  Thank you Dr. Joel for your reply.
I am very new to NWAs so, it might sound a _stupid question_.. but what will happen *if the pulse is much shorter than the Full sweep?* 

For e.g sweep time is 250microsec and the pulse width is 100 microsec(Pulse period 1000microsec). What happens in the NWA in such a condition? What makes the data capture invalid?

I want a favor from Dr. Joel or all other who read this post. Is there any specific book that will give me basic ideas about NWAs? A kind of crash course approach would suit me best. (I recently found a book from Dr. Joel in the library). Anything that you had as a reference/*used to gain/further your knowledge* would be highly appreciated.

Edited by: APL on Jan 25, 2015 7:27 PM

扫描时间是指完成扫描的时间。
它可能大约5-50毫秒！
如果在脉冲开始时触发扫描，脉冲将在扫描结束前消失。
但更常见的是脉冲测量的点触发，这里重要的是时间/点。
每点的时间约为1 / IFBW。
如果IF BW为100 kHz，则每点的时间为10 usec。
如果IF BW为10 kHz，则每点的时间为100 usec。
如果脉冲短于每点的时间，则IF测量将在某些时间没有信号，并且您将具有较差的迹线噪声。
这里关键的是控制时序，以便脉冲在触发之前开始，并在IF BW完成测量之后结束。
顺便说一句，仅供参考，但这是PNA上全自动脉冲测量应用程序。
包括一个内置的可选脉冲发生器。
在ENA上，您必须做一些工作才能使所有内容正确同步。
至于另一个问题，对于像脉冲这样的复杂测量，你也可以查看按键应用笔记（谷歌“PNA脉冲测量”）;
我的书中有几页关于放大器的脉冲RF测量。



     以下为原文

  Sweep time refers to the time to take the entire sweep.  It is likely on the order of 5-50 msec!  If you trigger a sweep when the pulse starts, the pulse will be gone way before the sweep finishes.

But much more common is point-trigger for pulsed measuremends, and here what is important is time/point.  Time per point is about 1/IFBW.

If the IF BW is 100 kHz, then the time per point is 10 usec. If the IF BW is 10 kHz, then the time per point is 100 usec.  If the pulse is shorter than the time per point, then the IF measurement will have some of the time with no signal and you will have poorer trace noise.  What is critical here is controlling the timing so the pulse starts just before the trigger, and ends just after the IF BW finishes the measurement.

BTW, just FYI but this is all fully automatic on the PNA with pulse measurement app. including a built-in optional pulse generator.  On the ENA you have to do a bit of work to get all everything properly synched.

As to the other question, for complicated measurements like pulse, you can also look to keysight application notes (google "PNA pulse measurement"); There are a few pages in my book on pulsed RF measurements of amplifiers.

谢谢Joel博士我正在阅读你的书，以便更好地理解脉冲测量。
从你的答案和我有限的理解。
*（如果我错了，请纠正我）*点+测量+的缺点可能是由IFBW造成的。
测量IFBW> 1 /脉冲宽度。
但是，随着我们继续增加IFBW值，扫描时间变得非常慢......而且，NWA可能不包含适用于特定脉冲测量的IFBW ......扫描时*最有影响的缺点是什么？
和* on point *技术应用于脉冲测量？
另外，你说典型的扫描时间是5-50毫秒！
什么控制扫描时间？
（我知道这是一个愚蠢的问题）_为什么它不能更快​​？_



     以下为原文

  Thank you Dr. Joel

I am reading your book to better understand pulsed measurement. From your answers and based on my limited understanding. *(Please correct me if I am wrong)* The disadvantage of +on point+ measurement can be caused by IFBW. For measurement IFBW>1/pulse width. But, as we go on increasing IFBW value, the sweep time becomes very slow... Also, NWA may not contain IFBWs that will be suitable for a particular pulsed measurement...

What would be the most influential disadvantage of *on sweep* and *on point* techniques when applied to pulsed measurements?

Also, you said that typical sweep times are 5-50msec! What controls the sweep time? (I know it is a stupid question) _Why cannot it be faster?_

> {quote：title = APL写道：} {quote}>我正在读你的书，以便更好地理解脉冲测量。
从你的答案和我有限的理解。
*（如果我错了，请纠正我）*总是乐意：>点+测量+的缺点可能是由IFBW造成的。
测量IFBW> 1 /脉冲宽度。
但是，随着我们继续增加IFBW值，扫描时间变得非常慢......而且，NWA可能不包含适合特定脉冲测量的IFBW ......不太正确。
当您-REDUCE- IFBW时，每点的时间会上升（增加），因此无法测量较短宽度的脉冲。
或者以数学方式说，IF带宽的数据采集时间是t = 1 /（IFBW）。
对于短脉冲，您必须具有宽的IF BW，但PNA窄带脉冲模式除外（稍后将详细介绍）>>当应用于脉冲时，*扫描*和*点*技术的最大影响是什么？
测量？
>>另外，你说典型的扫描时间是5-50毫秒！
什么控制扫描时间？
（我知道这是一个愚蠢的问题）_为什么它不能更快​​？_事实上，没有人使用“扫描”脉冲测量，因为扫描时间太慢，很少有任何情况下脉冲准时足够长
整个数据扫描。
如果扫描中有N个测量点，则扫描时间约为N /（IFBW）。
因此，扫描模式总是比点模式差，并且大多数时间脉冲太短而不能使用每个脉冲一次扫描。
在PNA / PNA-X中，我们有一种称为“窄带脉冲模式”的特殊模式，它允许用任何IFBW测量脉冲。
细节非常复杂，需要PNA上的附加选项，但它可以支持脉冲测量，脉冲宽度约为20纳秒。
BTW：你需要什么样的脉冲宽度，以及你测量的是什么样的DUT。



     以下为原文

  > {quote:title=APL wrote:}{quote}
> I am reading your book to better understand pulsed measurement. From your answers and based on my limited understanding. *(Please correct me if I am wrong)* 
Always happy to:
>The disadvantage of +on point+ measurement can be caused by IFBW. For measurement IFBW>1/pulse width. But, as we go on increasing IFBW value, the sweep time becomes very slow... Also, NWA may not contain IFBWs that will be suitable for a particular pulsed measurement...
Not quite correct.  As you -REDUCE- the IFBW, the time per point goes UP (increases) and so it cannot measure shorter width pulses. Or to say it mathematically, the data-acquistion time for an IF bandwidth is t~=1/(IFBW).   For short pulses you must have wide IF BW, with a key exception of the PNA narrowband pulse mode (More on that later)

> 
> What would be the most influential disadvantage of *on sweep* and *on point* techniques when applied to pulsed measurements?
> 
> Also, you said that typical sweep times are 5-50msec! What controls the sweep time? (I know it is a stupid question) _Why cannot it be faster?_

In fact, no one uses "on sweep" pulsed measurements because the sweep time is so slow there are seldom any cases where the pulse on time is long enough to get an entire sweep of data.

If you have N measurment points in a sweep, the sweep time is approximately N/(IFBW).  Thus sweep mode is always worse than point mode, and most of the time pulses are too short to use one sweep per pulse.

In the PNA/PNA-X we have a special mode that is called "narrowband pulse mode" and it allows measurements of pulses with any IFBW.  The details are really complicated, and requires an additional option on the PNA, but it can support pulse measurements down to pulse widths around 20 nano-seconds.   

BTW: what is the pulse width you need, and what kind of DUT are you measureing.

DUT是一种特殊类型的* +微波谐振器天线+ *。
将天线插入脉冲（0.4-20kHz）的等离子体中。
来自NWA的微波激发天线，并从ENA屏幕读取共振频率峰值。
共振信号取决于等离子体电子密度（电子密度取决于所馈送的功率）。
在脉冲等离子体中，共振频率随时间变化（由于等离子体电子密度的变化）。
所以，我需要将测量与脉冲频率同步。
* _我想测量时间分辨共振频率._ *在开始时我没有注意到脉冲技术，所以我使用了正常模式，即扫描。
我可以获得满意的结果，高达400Hz脉冲频率，占空比为13％。（脉冲宽度= 325微秒，脉冲周期2500微秒）。
但是，从病房1kHz开始......特别是随着占空比增加（脉冲导通时间越长），所获得的数据不符合我们的理论预测。
所以我们放弃了 - 扫一扫 - 然后继续说道。
现在，我只使用'on point'技术，但我需要解释为什么'on sweep'以某种方式适用于较低频率但不适用于较高频率。
我一直在寻找这个地方..到目前为止它几乎是一个疯狂的追逐！编辑：APL于2015年2月9日12:08 AM



     以下为原文

  The DUT is a special type of *+microwave resonator antenna+*. The antenna is inserted in a plasma that is pulsed from (0.4-20kHz). The microwave from NWA excites the antenna and resonance frequency peak is read from the ENA screen. The resonance signal depends on plasma electron density (electron density depends on the power being fed).

In a pulsed plasma the resonance frequency changes with time(owing to changing plasma electron density). So, I need to synchronize measurement with pulsed frequency. *_I want to measure time-resolved resonance frequency._*

In the beginning I was oblivious of  point on pulse technique, so I used the normal mode i.e. on sweep. I could get satisfactory results upto 400Hz pulse frequency with duty 13%.( pulse width = 325 microsec and pulse period 2500 microsec). But, from 1kHz on wards.. especially with increasing duty(longer pulse on time) the data obtained didn't adhere to our theoretical prediction. so we discarded --on sweep-- and went for on point.

Now, I only use 'on point' technique, but I need to explain why 'on sweep' somehow worked for lower frequencies but not for higher ones. I have been looking for this in several places.. so far it has almost been a wild goose chase!

Edited by: APL on Feb 9, 2015 12:08 AM

Joel博士我有一个问题（开始一个新的主题 - 没有人回复）。
使用NWA内置触发延迟和使用外部设备延迟输入信号（馈送到外部触发端口）有什么区别？



     以下为原文

  Dr Joel

I have a question (started a new thread-no reply from anyone yet). What is the difference between using NWA inbuilt trigger delay and using an external device to delay the incoming signal(to be fed to External trigger port)?

以下是从PNAHelp主题到非常基本的交互式冲击波演示的链接：http：//na.support.keysight.com/pna/NaBasics/network_analysis_basics.htm它已经陈旧但大部分仍然相关。



     以下为原文

  Here is a link from a PNAHelp topic to a very basic, interactive shockwave presentation : 
http://na.support.keysight.com/pna/NaBasics/network_analysis_basics.htm

It is old but most of it is still relevant.

谢谢你的flash视频。
这真的很有帮助。
我只是想知道，为什么NWA公司不会像操作软件那样进行一些关于NGA操作/基本使用的在线辅导课程！
这对我这样的人来说真的很有帮助！



     以下为原文

  Thank you for the flash video. It was really helpful. I was just wondering, why NWA companies don't conduct some online tutorial sessions on operation/basic use of NWAs, just like software guys do!

It would be really be helpful(for people like me!).