> {quote：title = GT_Test写道：} {quote}> *问题1：* S21代替DUT在校准后不平坦（> 1.5 dB）>我首先在端口上使用功率计/传感器进行功率校准
1（-25dBm至0 dBm，步长为1 dB）。
>然后我用eCal模块（N4691-60004）更换我的DUT并进行全2端口S-param校准（端口功率为-10 dBm）。
>接下来我把_thru放入我的DUT_然后看S21（dB）并且没有得到我期望的平坦度。
我尽量减少测试电缆的移动，并小心所有连接。
>如果我做了响应，那么我可以得到平坦度我有不同的IFBW，平均值，初始端口功率电平，测试电缆，测试耦合器，具有类似的结果。
>因为您将30 dB耦合系数外部耦合器的耦合臂连接到端口2上的测试端口，您基本上在端口2路径中添加了30 dB衰减器。
当您使用-10 dBm端口电源执行完整的2端口S参数ecal时，连接到端口2的ecal端口的功率电平为-40 dBm，并且在校准期间接收器B测量的任何反射信号将为
-70 dBm或更低。
如果你想在这种配置中获得更好的2端口校准，你必须解除端口功率并将端口2的功率设置为尽可能高而不会不平衡并减少IFBW（和/或增加平均值
在cal）期间。
您的另一个选择是只执行EnhancedResponse cal，这将允许您测量S11和S21，但不能测量S22。
这是大多数具有高功率设置的人通常所做的事情，因为您将测量的S22基本上是30 dB耦合臂的匹配，而不是放大器输出端口的匹配。
但是，如果您真的想要一个良好的完整2端口校准，那么您将需要设置一个双外部反射计。
下面是PNA-X上一个这样的设置的示例框图，但这个概念也适用于您的E8363C PNA。
在下面的示例中（来自高功率应用笔记1408-19），您可以忽略输入外部放大器，但如果您想要进行完整的2端口校正，则输出双反射计设置就是您所需要的。
！HTTPS：//dl.dropboxusercontent.com/u/87949221/Forum/Thread_119233/HighPowerTestset.png！
以上来自于谷歌翻译
以下为原文
> {quote:title=GT_Test wrote:}{quote}
> *Issue1:* S21 with thru in place of DUT is not flat (>1.5 dB) after calibration
> I first conduct a power calibration with my power meter/sensor on Port 1 (-25dBm to 0 dBm in 1 dB steps).
> Then I replace my DUT with the eCal module (N4691-60004) and do a full 2 port S-param calibration (at port power of -10 dBm).
> Next I put a _thru in for my DUT_ and look at S21 (dB) and do not get the flatness I would expect. I try to minimize test cable movement and am careful with all connections.
> If I do a Response Thru cal then I can get flatness I have varied IFBW, Averaging, Initial port power level, test cables, test coupler, with similar results.
>
because you are connecting the coupled arm of your 30 dB coupling factor external coupler to the test port on port 2, you are basically adding a 30 dB attenuator in the port 2 path. when you do your full 2-port S-Parameter ecal with -10 dBm port power, the ecal port that is connected to port 2 is seeing a power level of -40 dBm and any reflected signal measured by receiver B during the cal will be -70 dBm or lower. if you want to get a better 2-port cal in this configuration, you'll have to uncouple the port powers and set the port 2 power as high as it can go without going unleveled and reducing the IFBW (and/or increasing the averaging during the cal). Your other option is to just perform an EnhancedResponse cal, which will allow you to measure S11 and S21, but not S22. this is what most people with high power setups usually do, since the S22 that you will measure will be essentially the match of the 30 dB coupled arm and not the match of your amplifier's output port.
However, if you really want a good full 2-port calibration, then you will need to setup a dual external reflectometer. below is an example block diagram of one such setup on a PNA-X, but the concept applies to your E8363C PNA as well.
in the example below (from the high power app note 1408-19) you can ignore the input external amplifier, but the output dual reflectometer setup is what you need if you want to do a full 2-port correction.
!https://dl.dropboxusercontent.com/u/87949221/Forum/Thread_119233/HighPowerTestset.png!
> {quote：title = GT_Test写道：} {quote}> *问题1：* S21代替DUT在校准后不平坦（> 1.5 dB）>我首先在端口上使用功率计/传感器进行功率校准
1（-25dBm至0 dBm，步长为1 dB）。
>然后我用eCal模块（N4691-60004）更换我的DUT并进行全2端口S-param校准（端口功率为-10 dBm）。
>接下来我把_thru放入我的DUT_然后看S21（dB）并且没有得到我期望的平坦度。
我尽量减少测试电缆的移动，并小心所有连接。
>如果我做了响应，那么我可以得到平坦度我有不同的IFBW，平均值，初始端口功率电平，测试电缆，测试耦合器，具有类似的结果。
>因为您将30 dB耦合系数外部耦合器的耦合臂连接到端口2上的测试端口，您基本上在端口2路径中添加了30 dB衰减器。
当您使用-10 dBm端口电源执行完整的2端口S参数ecal时，连接到端口2的ecal端口的功率电平为-40 dBm，并且在校准期间接收器B测量的任何反射信号将为
-70 dBm或更低。
如果你想在这种配置中获得更好的2端口校准，你必须解除端口功率并将端口2的功率设置为尽可能高而不会不平衡并减少IFBW（和/或增加平均值
在cal）期间。
您的另一个选择是只执行EnhancedResponse cal，这将允许您测量S11和S21，但不能测量S22。
这是大多数具有高功率设置的人通常所做的事情，因为您将测量的S22基本上是30 dB耦合臂的匹配，而不是放大器输出端口的匹配。
但是，如果您真的想要一个良好的完整2端口校准，那么您将需要设置一个双外部反射计。
下面是PNA-X上一个这样的设置的示例框图，但这个概念也适用于您的E8363C PNA。
在下面的示例中（来自高功率应用笔记1408-19），您可以忽略输入外部放大器，但如果您想要进行完整的2端口校正，则输出双反射计设置就是您所需要的。
！HTTPS：//dl.dropboxusercontent.com/u/87949221/Forum/Thread_119233/HighPowerTestset.png！
以上来自于谷歌翻译
以下为原文
> {quote:title=GT_Test wrote:}{quote}
> *Issue1:* S21 with thru in place of DUT is not flat (>1.5 dB) after calibration
> I first conduct a power calibration with my power meter/sensor on Port 1 (-25dBm to 0 dBm in 1 dB steps).
> Then I replace my DUT with the eCal module (N4691-60004) and do a full 2 port S-param calibration (at port power of -10 dBm).
> Next I put a _thru in for my DUT_ and look at S21 (dB) and do not get the flatness I would expect. I try to minimize test cable movement and am careful with all connections.
> If I do a Response Thru cal then I can get flatness I have varied IFBW, Averaging, Initial port power level, test cables, test coupler, with similar results.
>
because you are connecting the coupled arm of your 30 dB coupling factor external coupler to the test port on port 2, you are basically adding a 30 dB attenuator in the port 2 path. when you do your full 2-port S-Parameter ecal with -10 dBm port power, the ecal port that is connected to port 2 is seeing a power level of -40 dBm and any reflected signal measured by receiver B during the cal will be -70 dBm or lower. if you want to get a better 2-port cal in this configuration, you'll have to uncouple the port powers and set the port 2 power as high as it can go without going unleveled and reducing the IFBW (and/or increasing the averaging during the cal). Your other option is to just perform an EnhancedResponse cal, which will allow you to measure S11 and S21, but not S22. this is what most people with high power setups usually do, since the S22 that you will measure will be essentially the match of the 30 dB coupled arm and not the match of your amplifier's output port.
However, if you really want a good full 2-port calibration, then you will need to setup a dual external reflectometer. below is an example block diagram of one such setup on a PNA-X, but the concept applies to your E8363C PNA as well.
in the example below (from the high power app note 1408-19) you can ignore the input external amplifier, but the output dual reflectometer setup is what you need if you want to do a full 2-port correction.
!https://dl.dropboxusercontent.com/u/87949221/Forum/Thread_119233/HighPowerTestset.png!
举报
HP_PNA setup.png66.9 KB
