怎么使用端口扩展参考平面？

不同的分析仪具有不同的功能来处理这个问题
你的VNA的型号和固件版本是什么？



     以下为原文

  Different analzyers have different functionaliy to handle this.  What's your model and firmware rev of VNA?

你好dr.joel，我有一个f8 rev 9.42的e8362c，还有一个Pna-x Reds Mahesh



     以下为原文

  Hello dr.joel,

I have an e8362c with fw rev 9.42 and also a  Pna-x

Reds
Mahesh

您好，有人可以让我知道有关端口扩展的查询的答案。
Rgds Mahesh



     以下为原文

  Hello,

Could somebody please let me know the answer for the query regarding port extension.

Rgds
Mahesh

PNA和PNA-X分析仪中的端口扩展功能可以校正电缆或夹具的损耗和延迟，但它不能校正DUT参考平面上的任何失配误差。
因此，如果您能够忍受这种不确定性，那么您当然可以使用PNA的自动端口扩展功能。
但是，在每个开关矩阵端口的末尾，您还有另一种选择。
如果您在PNA端口和待测参考平面之间的开关矩阵中表征（即测量4个s参数）每个单独的路径并将s参数存储在S2P文件中，您还可以使用PNA的2端口de-
通过夹具菜单访问的嵌入功能。
如果每条路径的损耗小于20 dB，则此选项可以正常工作。
但是，如果损耗大于20 dB，即使端口扩展功能也不会给您带来好的结果。



     以下为原文

  the port extension feature in the PNA and PNA-X analyzers can correct for the loss and delay of the cables or fixtures, but it does not correct for any mismatch errors at the DUT reference plane.  So, if you can live with that uncertainty, then you can certainly use the Automatic Port Extension feature of the PNA.  However, you do have another option short of calibrating at the end of each switch matrix port.  if you characterize (i.e measure the 4 s-parameters) each individual path in the switch matrix between the PNA port and the dut reference plane and store the s-parameters in an S2P file, you can also use the PNA's 2-port de-embedding feature accessed through the fixturing menu.  this option works great if the loss through each path is less than 20 dB.  however if you have loss greater than 20 dB, even the port extension feature will not give you good results.

一种方法是使用1端口表征和去嵌入;
类似的方法是使用重复的1端口校准和未知的通过;
然而，第三种（也许是首选的）是使用QSOLT方法。
通过执行单路径引导功率校准（具有S参数），可以校准用于S参数的开关矩阵的每个路径和功率测量。
对于每个其他端口或路径，仅需要1端口校准，并且路径差异可以是去嵌入以给出每个路径的最终结果。
该方法为每条路径生成一个S2P文件，即第一个完全校准路径和每个其他端口路径之间的差异;
以这种方式，它与Daras讨论的非常相似。
或者，如果您可以为路径定义直通，则可以使用QSOLT方法：执行1端口校准并通过单端口对之间定义。
图1A和2A（字母表示swtich路径）。
然后从1A连接到2B并加载1端口校准并通过直通步骤完成校准（必须以编程方式完成）。
然后QSOLT从2B到1B，然后再从1B到2C，然后是2C到1C等。因此，你有一个1端口校准和2N-1步骤。
也可以进行其他可能的组合。



     以下为原文

  One method is to use 1-port characterization and de-embedding; a similar approach is to use repeated 1 port calibrations and unknown thru; yet a third (and maybe prefered) is to use a QSOLT approach.

It is possible to calibrate every path of a switch matrix for S-parameters and a power measurements by doing a single path Guided Power Calibration (with S-parameters).  The for every other port or path, only a 1 port calibration is needed, and the path difference can be de-embedding to give the final result for each path.  This method yields an S2P file for each path, that is the difference between the first full-calibration path and every other port path; in this way it is quite similar to what Daras discusses.

Alternatively, if you can define a thru for the paths, you can use a QSOLT approach: do a 1 port cal and defined through between one-port pair. 1A and 2A (letters indicate swtich paths).  Then connect through from 1A to 2B and load the 1-port cal and complete the cal with the thru step (must be done programatically).  Then QSOLT from 2B to 1B, then again from 1B to 2C, then 2C to 1C, etc.

Thus, you have one 1-port calibration and 2N-1 thru steps.

Other possible combinations can be done as well.