未知直通测量上的波纹

> {quote：title = bar28写道：} {quote}>我用我的E5071C ENA使用85054D校准套件和IFBW 10Hz完成了Uknown Thru校准（两个testport电缆的末端公连接器）。
通过组件我使用calkit f-f无槽适配器，我之前测量的电延迟。
为什么传输侧S21在测量校准后会出现这样的纹波（见附件）？
Testport电缆固定良好，因此不能从那里产生波纹。
S11和S22的结果看起来不错。
我重复了整个校准并检查了所有连接器，但结果完全相同。
毫无疑问，是德科技的某个人会发表评论，但看着这一点，我看不出有任何错误。
每个除法为0.01 dB。
你得到的更糟糕的是0.02 dB的峰 - 峰纹波，所以我会很高兴。
戴夫



     以下为原文

  > {quote:title=bar28 wrote:}{quote}
> I done Uknown Thru calibration (both testport cable's endings male connectors) with my E5071C ENA using 85054D calibration kit and  IFBW 10Hz. Thru component i used calkit f-f slotless adapter which electrical delay i measured earlier. Why transmission side S21 has such a ripple (see attachement) after measurement calibration? Testport cables are well fixed so ripple can't come from there. S11 and S22 results looks fine. I repeated whole calibration and checked all connectors, but results were totally same.


No doubt someone from Keysight will comment, but looking at that, I don't see anything too wrong. Each division is 0.01 dB. The very worse you got is a 0.02 dB peak-peak ripple, so I would be happy with that. 

Dave

dr_kirby是对的。
结果实际上比预期好一点。
对于您使用的ENA和calkit，对不确定性的简单估计大约为+ -0.14 dB p-p;
您可以使用Keysight不确定性计算器来查找使用ecal的值，大约为0.11 dB p-p（85054未在不确定性计算器中列出，但负载匹配稍差，因此如上所述稍差）。
具有最佳calkit的最佳计量选项PNA将在通过时获得0.08 dB p-p误差。
那是因为那些频率下N型的最佳负载并不是那么好。
编辑:(实际上，列出的是85045，不是ENA的默认值;不确定性计算器给出了一个略大的数字，对于ENA约为0.16 dB p-p）编辑：Dr_joel于2015年6月12日上午8:25



     以下为原文

  dr_kirby is correct.  The results are actually quite a bit better than expected.

For the ENA and calkit that you use a simple estimate of the uncertainty is on the order of +-0.14 dB p-p; and you can use the Keysight uncertainty calculator to find the value for using ecal on the order of 0.11 dB p-p (the 85054 isn't listed in the uncertainty caculator but is slightly worse load match so slightly worse as indicated above).  The best metrology option PNA with the best calkit will get you 0.08 dB p-p error on a thru.  That's because the best loads in type N at those frequencies just aren't that good.

edit: (Actually, the 85045 is listed, just not in defaults for ENA; the uncertainty calculator gives a slightly larger number around 0.16 dB p-p for ENA)

Edited by: Dr_joel on Jun 12, 2015 8:25 AM