如何测量耦合端口

如果你碰巧有一个4端口PNA，那么你可以简单地做一个3端口Smart Cal，其中两个端口连接器类型设置为Waveguide，并且为这些端口选择R11664A calkit，第三个端口设置为连接器类型
同轴耦合端口，您为该同轴连接器选择适当的calkit。
当您执行校准时，将提示您为每个端口提供适当的标准，但是您还需要有一个波导来同步适配器以用于未知的直通步骤（未知通过默认的直通方法）。
如果你只有2个PNA，那么你需要做2端口校准，其中1个端口设置为波导，另一个设置为同样如上所述。
你仍然需要一个波导来哄适配器用于直通步骤。
当您进行测量时，您需要使用波导负载端接未使用的波导端口。



     以下为原文

  If you happen to have a 4 port PNA, then you can simply do a 3 port Smart Cal where two of the port connector types are set to Waveguide and the R11664A calkit is selected for those ports and the 3rd port is set to the connector type of the coaxial coupled port and you select the appropriate calkit for that coaxial connector.  When you do the cal, the will prompt you for the appropriate standards for each of the ports, but you will also need to have a waveguide to coax adapter for the unknown thru steps (unknown thru is the default thru method).  

if you only have a 2 PNA, then you need to do 2-port cal where 1 port is set to waveguide and the other set to coax just as above.  you will still need a waveguide to coax adapter for the thru step.  when you do the measurement, you will need to terminate the unused waveguide port with a waveguide load.

daras，感谢您的快速回复。
PNA是一个2端口型号，所以我尝试了你的第二个选项。
R11664A校准套件具有固定负载的移动波导负载 - 我认为它在校准过程中以相同的方式使用吗？
此外，对于直通步骤，我通过适配器提示延迟。
我接受Smartcal的默认值吗？
我正在使用R281B WG /同轴适配器，但我没有关于延迟的信息。
再次感谢。



     以下为原文

  daras,

Thanks for the prompt reply.  The PNA is a 2 port model, so I tried your second option.  The R11664A cal kit has a moving waveguide load vice a fixed load - I assume it is used in the same manner during the cal process?  Also, for the thru step, I'm prompted for the delay thru the adapter. Do I accept the default value from Smartcal?  I'm using R281B WG/coax adapters, but I have no info on the delay. Thanks again.

是的 - 您应该能够接受被引导校准的计算延迟。
在使用波导适配器的某些情况下，模式反转会导致计算出的延迟显示为负数，但我认为该问题已在最新固件中得到解决。
如果延迟值看起来合理，那么它是正确的。
根据您运行的固件版本，您有一个名为Characterize Adapter Macro的实用程序或Response-> Cal-> Fixtures-> Cal Plane Manager下的菜单项。
使用这些工具中的任何一个，您可以使用适配器两端的两个1端口校准来测量适配器的S参数（包括波导到同轴适配器）。
如果您不确定适配器的延迟，可以先对其进行表征，然后查看结果的S21延迟。



     以下为原文

  Yes - you should be able to accept the calculated delay by the guided cal.  In some cases with waveguide adapters a mode inversion causes the calculated delay to show up as a negative number, but I think that problem has been solved in latest firmware.  if the delay value looks reasonable, then it is correct.  depending on the version of firmware you are running, you have either a utility called the Characterize Adaptor Macro or menu item under Response->Cal->Fixtures->Cal Plane Manager. with either of these tools, you can measure the S-Parameter of an adapter (including a waveguide to coax adapter) using two 1-port calibrations on either end of the adapter.  if you are not sure about your adapter's delay, you can characterize it first and look at the S21 delay of the results to be sure.

滑动可能不会给出特别好的隔离结果（它应该适用于耦合），因为滑动负载元件可能没有非常好的回波损耗。
这是因为滑动载荷通过测量不同偏移处的（可能）相同的不良载荷来获得其质量，以计算表示通过理想线滑动的载荷的不匹配的点的轨迹的中心。
一些滑动负载与20或26 dB回波损耗差不多，有些则要好得多。
因此，做一个1端口滑动负载波导cal，然后重新打开滑动负载并测量其在各种滑动位置的回波损耗，看看它的外观。



     以下为原文

  The sliding may not give a particularly good result for isolation (it should be fine for coupling) because the sliding load element may not have very good return loss.  That is because the sliding load gets its quality from measuring the (presumably) same bad load at different offsets to compute the center of a locus of points that represents the bad match of the load slid through an ideal line.   Some sliding loads are as bad as 20 or 26 dB return loss, some are much better.  So, do a 1 port sliding load waveguide cal and then put the sliding load back on and measure its return loss at various slide positions to see how it looks.