如何校准N5532A传感器模块

关于N5531S应用笔记+“使用Agilent N5531S测量接收器系统进行精确绝对和相对功率测量”+，我有一个问题。
我想更好地理解附录A，它着眼于绝对功率精度。
但我的问题是关于如何校准N5532A传感器模块。
现在在AN的第5页，它说N5532A作为一个单元进行校准，通过用负载端接RF输出端口并将该单元表征为+“......如果它是常规功率传感器”+。
那么，以这种方式校准N5532A是首选吗？
另一种方法是从设备上拆下功率传感器，单独校准然后重新组装。
我的猜测是不建议拆卸装置，但也许我错了。
附录A似乎另有说明。
第13和14页描述了不确定性分析所需的详细测量方程。
它们包括内部N5532A接口的不匹配项（例如，在分路器输出和功率传感器输入之间），如果功率传感器被独立地表征，则这将导致不确定性（否则这些接口的影响包括在复合校准因子中）。
因此，如果您能澄清校准N5532A的首选方法，我将不胜感激。
此外，如果有人能够对附录C中的倒数第二段有所启发，我将不胜感激。
它说+“......这种情况相当于它们被校准的情况。因此它们都是同时最佳精确配置的双电阻分离器......”+这个短语似乎说校准
功率传感器的原位完成。
但我不明白为什么如果选择了一个分隔线，情况会有所不同。
提前致谢 ：-）



     以下为原文

  I have a question about the N5531S Application Note +"Accurate Absolute and Relative Power Measurements Using the Agilent N5531S  Measuring Receiver System"+.

I would like to better understand Appendix A, which looks at the absolute power accuracy. But my question is really about how the N5532A sensor module is calibrated.  

Now on page 5 of the AN, it says that the N5532A is calibrated as a unit, by terminating the RF output port with a load and characterizing the unit as +"... if it were a regular power sensor"+.

So, is calibration of the N5532A in this way preferred? The alternative would be to remove the power sensor from the unit, calibrate it separately and then reassemble everything.

My guess is that disassembling the unit is not recommended, but perhaps I am wrong. Appendix A appears to suggest otherwise. Pages 13 and 14 describe the detailed measurement equations required for an uncertainty analysis. They include mismatch terms at internal N5532A interfaces (e.g., between splitter output and power sensor input) that would only contribute to uncertainty if the power sensor was characterized independently (otherwise the effect of these interfaces is included in the composite calibration factor).

So, I would be grateful if you could clarify the preferred method of calibrating the N5532A.

Also, if anyone is able to shine some light on the second to last paragraph in Appendix C I would be grateful. It says there

+"... This situation is equivalent to the situation under which they were calibrated. Thus they are both, simultaneously optimally accurate as configured with a two-resistor splitter. ..."+

This phrase seems to say that calibration of the power sensor is done in-situ. But I don't see why the situation would be any different if a divider had been chosen.

Thanks in advance  :-)