引用: 脑洞大赛9 发表于 2019-1-22 10:44
首先，确保R2在端口1处定义源。我不确定ENA的确切语法;
对于PNA，它是R2_1（但R2的默认值是R2_2，意味着在反向扫描期间测量R2）。
其次，请注意，如果端口1通过电缆连接到端口2，则信号进入端口2，从内部源反弹并重新反射出......这由R2_1测量。
感谢Dr_joel的回答..我联系了安捷伦支持人员并了解问题所在。
我的算法不起作用，因为我下载和处理的数据是由仪器以数学方式操作的。
对于我的NA型号E5071C，绝对测量在数学上被校正，因为它具有更好的“虚拟方向性”，这是“帮助用户”。
不幸的是，当通过另一个端口提供激励时，这种校正不适用于每个端口的参考接收器，例如R2（1）由于这个原因，数据对于执行依赖于8项误差模型的外部校准是无用的（例如：TRL）
校准）。
对于在12项误差模型（例如TOSM_）上进行中继的所有校准都没有问题，因为它们不需要Rn（m）的信息。
要在依赖于8term错误模型的算法中执行数据的外部使用，必须在关闭“虚拟桥”和其他校正之前使Rn（m）数据变得有用。
http://ena.tm.agilent.com/e5071c/manuals/webhelp/eng/遗憾的是，手册中描述的程序不起作用，因为仪器无法识别这些SCPI命令。
我再次询问并等待安捷伦的支持。
如果您对此信息感兴趣，请与我们联系。
其他一些仪器如E5071C 14和20 GHz没有这个问题，因为它们没有数学校正，我想这也是高端仪器的情况，如PNA和PNAX。
看你的答案我反对“第二”点不同意，因为当源在端口1时，端口2的溢出关闭且端口匹配得很好......因此很少有功率应该返回并从R2测量（1
）。
我们在R2（1）中有一个重要的衡量标准，因为“上面提到的”方向性差，使得这个接收器对“进入”已经由B（1）接收机测量的端口的波敏感。
这也发生在端口1中，但仪器校正A（1）和R1（1）的测量值以改善差的方向性但不正确的R2（1）。
感谢干杯亚历山德罗的“第三”信息
以上来自于谷歌翻译
以下为原文
Thank you Dr_joel for your answer..
I contacted the Agilent support and understood what the problem was.
My algorithm is not working because the data I downloaded and process are mathematically manipulated by the instrument.
For my NA model, E5071C, the absolute measurement are mathematically corrected tho have a better "virtual directivity" and this is to "help the user".
Unfortunately this correction doesn't apply to reference receiver of each port when the stimulus is provided through another port
eg R2(1)
For this reasons the data are not useful to perform external calibration that rely on the 8 term error model ( eg: TRL calibration).
No problem for all those calibrations that relay on the 12 term error model ( eg TOSM_) because they no need the information of Rn(m).
To perform external use of the data in algorithms that rely in the 8term error model you must before turn OFF the "vitual bridge" and other corrections so that also the Rn(m) data became useful.
http://ena.tm.agilent.com/e5071c/manuals/webhelp/eng/
Unfortunately the procedure described in the manual doesn't work because the instrument doesn't recognize those SCPI commands.
I'm again asking and waiting from Agilent support on that. Let me know if you are interested on this information.
Some other instrument as E5071C 14 and 20 GHz have not this problem because they have no mathematical correction and I suppose this is also the case of high end level instrument as PNA and PNAX.
Looking to your answer I anyway disagree with "second" point because when the source is on port 1 the surce in port 2 is off and the port is quite well matched..so that few power should come back and get measured from R2(1). We have a significant measure in R2(1) because of the "above mentioned" poor directivity that make this receiver sensitive also to the wave "entering" the port that is already measured by the B(1) receiver.
This happens also in port 1 but the instrument correct the measurement of A(1) and R1(1) to improve the poor directivity but not correct R2(1).
Thank you for the "third" information
Cheers
Alessandro
引用: 脑洞大赛9 发表于 2019-1-22 10:44
首先，确保R2在端口1处定义源。我不确定ENA的确切语法;
对于PNA，它是R2_1（但R2的默认值是R2_2，意味着在反向扫描期间测量R2）。
其次，请注意，如果端口1通过电缆连接到端口2，则信号进入端口2，从内部源反弹并重新反射出......这由R2_1测量。
感谢Dr_joel的回答..我联系了安捷伦支持人员并了解问题所在。
我的算法不起作用，因为我下载和处理的数据是由仪器以数学方式操作的。
对于我的NA型号E5071C，绝对测量在数学上被校正，因为它具有更好的“虚拟方向性”，这是“帮助用户”。
不幸的是，当通过另一个端口提供激励时，这种校正不适用于每个端口的参考接收器，例如R2（1）由于这个原因，数据对于执行依赖于8项误差模型的外部校准是无用的（例如：TRL）
校准）。
对于在12项误差模型（例如TOSM_）上进行中继的所有校准都没有问题，因为它们不需要Rn（m）的信息。
要在依赖于8term错误模型的算法中执行数据的外部使用，必须在关闭“虚拟桥”和其他校正之前使Rn（m）数据变得有用。
http://ena.tm.agilent.com/e5071c/manuals/webhelp/eng/遗憾的是，手册中描述的程序不起作用，因为仪器无法识别这些SCPI命令。
我再次询问并等待安捷伦的支持。
如果您对此信息感兴趣，请与我们联系。
其他一些仪器如E5071C 14和20 GHz没有这个问题，因为它们没有数学校正，我想这也是高端仪器的情况，如PNA和PNAX。
看你的答案我反对“第二”点不同意，因为当源在端口1时，端口2的溢出关闭且端口匹配得很好......因此很少有功率应该返回并从R2测量（1
）。
我们在R2（1）中有一个重要的衡量标准，因为“上面提到的”方向性差，使得这个接收器对“进入”已经由B（1）接收机测量的端口的波敏感。
这也发生在端口1中，但仪器校正A（1）和R1（1）的测量值以改善差的方向性但不正确的R2（1）。
感谢干杯亚历山德罗的“第三”信息
以上来自于谷歌翻译
以下为原文
Thank you Dr_joel for your answer..
I contacted the Agilent support and understood what the problem was.
My algorithm is not working because the data I downloaded and process are mathematically manipulated by the instrument.
For my NA model, E5071C, the absolute measurement are mathematically corrected tho have a better "virtual directivity" and this is to "help the user".
Unfortunately this correction doesn't apply to reference receiver of each port when the stimulus is provided through another port
eg R2(1)
For this reasons the data are not useful to perform external calibration that rely on the 8 term error model ( eg: TRL calibration).
No problem for all those calibrations that relay on the 12 term error model ( eg TOSM_) because they no need the information of Rn(m).
To perform external use of the data in algorithms that rely in the 8term error model you must before turn OFF the "vitual bridge" and other corrections so that also the Rn(m) data became useful.
http://ena.tm.agilent.com/e5071c/manuals/webhelp/eng/
Unfortunately the procedure described in the manual doesn't work because the instrument doesn't recognize those SCPI commands.
I'm again asking and waiting from Agilent support on that. Let me know if you are interested on this information.
Some other instrument as E5071C 14 and 20 GHz have not this problem because they have no mathematical correction and I suppose this is also the case of high end level instrument as PNA and PNAX.
Looking to your answer I anyway disagree with "second" point because when the source is on port 1 the surce in port 2 is off and the port is quite well matched..so that few power should come back and get measured from R2(1). We have a significant measure in R2(1) because of the "above mentioned" poor directivity that make this receiver sensitive also to the wave "entering" the port that is already measured by the B(1) receiver.
This happens also in port 1 but the instrument correct the measurement of A(1) and R1(1) to improve the poor directivity but not correct R2(1).
Thank you for the "third" information
Cheers
Alessandro
举报
