可以使用时域门控来获得ECal校准的校准残差值吗

是。
添加航空公司和固定负载。
然后在航空公司的起点附近登机并查看频域中的响应。
这假设航空公司明显优于负载（在阻抗方面）并且优于Ecal。
可以直接使用负载，但通常Ecal比负载更好，所以你最终只是测量负载中的误差（当然，它实际上是负载和Ecal的残余方向性之间的差异，但是
如果负载明显比Ecal差，则它的误差占主导地位）。



     以下为原文

  Yes. Add the airline and a fixed load. Then gate around the start of the airline and look at the response in the frequency domain.  This presumes that the airline is significantly better than the load (in terms of impedance) and better than the Ecal.  

One could use the load directly, but usually the Ecal is better than the load, so that you end up just measuring the error in the load (of course, it is really the difference between the load and the residual directivity of the Ecal, but if the load is significantly worse than the Ecal, it's error dominates).

谢谢。
顺便说一句，你的书很出色，是迄今为止我见过的最好的这本书。



     以下为原文

  Thanks.  By the way, your book is outstanding, the best I've ever seen by far for this field.

ECAL在安捷伦工厂根据主要校准标准进行校准，并具有计量实验室航空公司的准确性。
因此，使用标准航空公司技术提取ECal残余误差是不够的。
验证工具包中的典型航空公司不够准确。
实际上，EC已被用于验证商用机械校准套件，而不是相反。
肯



     以下为原文

  ECals are calibrated at the Agilent factory based on primary calibration standards with metrology lab airline accuracy. Therefore, extraction of ECal residual errors using standard airline technique is not adequate. Typical airlines in verification kits are not accurate enough.  Actually, ECals have been used to verify commercially available mechanical cal kits, not the other way around.

Ken

我们需要关联使用不同开关矩阵构建的不同测试站。
我们希望使用剩余误差值作为度量来帮助我们理解校准的质量，因此测量能够在每个站点上产生。
有没有办法用ECal提取错误条款，还是应该恢复到这个实验的机械标准？



     以下为原文

  We need to correlate different test stations that are built with different switch matrixes.  We wanted to use the use the residual error values as a metric to help us understand the quality of calibration and hence measurement were able to produce on each station.  Is there a way to extract the error terms with ECal or should I revert to mechanical standards for this experiment?

是。
事实上，这是最好的方式，毫无疑问。
Ecal优于滑动负载校准，并且优于TRL，除非使用“短线”无珠度量计连接器和线路。
这很容易。
执行和ecal，打开Cal：ManageCals：Calset View以直接查看错误术语。
这些将包括VNA，电缆和开关直到参考校准平面的所有误差。
对于给定的一组残余误差（来自Ecal），可以直接计算任何系统的有效不确定性。
S21的数学是（使其非常简单）^ 1 ^：ResdiualSM_Ecal *（SMraw + LMraw）* 1.41;
是的，还有十几个其他术语，但它们并不多，这主要是你需要关注的。
如果DUT的S11和S22非常大，您也可以添加S11 * SM + S22 * LM术语Rss'd。
^ 1 ^参考：http：//www.wiley.com/WileyCDA/WileyTitle/productCd-1119979552.html



     以下为原文

  Yes. In fact this is the best way, no question.  Ecal is better than sliding load calibration, and better than TRL except if "short-line" beadless metrology connectors and lines are used.

It is easy. Do and ecal, turn on the Cal:ManageCals:Calset View to see the error terms directly.  These will include all the errors from the VNA, cables, and switches up to the reference calibration plane.   With a given set of residual errors (from the Ecal) the effective uncertainty from any system can be computed directly.

The math for S21 is (to make it very simple)^1^:  ResdiualSM_Ecal*(SMraw+LMraw)*1.41;  Yes, there are a dozen other terms, but they don't amount to much, this is primarily what you need to be concerned about.  If the S11 and S22 of the DUT is quite large, you can add S11*SM+S22*LM terms Rss'd in as well.

^1^Reference: http://www.wiley.com/WileyCDA/WileyTitle/productCd-1119979552.html

如果要关联不同站点的测量值，则统计测量比较可能是最佳方法。
您可以使用相同的ECal来校准每个站点，而不是提取残留错误，而不是比较不同设备的测量值，例如验证工具包中的测量值。
您需要做我们称之为可重复性和可重复性的研究。
对此的一个很好的参考是NIST的工程统计手册，第2章 - 测量过程表征。
http://www.itl.nist.gov/div898/handbook/这是我们用来关联全球关键测试系统的方法。
由于测量是矢量，因此应使用矢量统计分析。
肯



     以下为原文

  If you want to  correlate measurements of different stations, a statistical measurement comparison is probably the best way.  Instead of extracting residual errors, you can compare measurements of different devices, such as those in a verification kit, using the same ECal to cal each station.  You will need to do what we call a repeatabilty and reproducibility study. A good reference on this is NIST's Engineering Statistics Handbook, chapter 2 - Measurement Process Characterization.
http://www.itl.nist.gov/div898/handbook/

This is the method we use to correlate critical test systems around the world.  Since the measurements are vector quantities, vector statistical analysis should be used.

Ken