不良校准装置对我的测量有什么影响？

嘿，谢谢drkirkby的回应。
所以实际上我简化了上面的场景。
实际发生的是，这个工具包位于我们位于不同位置的另一个实验室中，他们通过电子邮件向我发送了问题，因此我没有可用的数据，只是上面的描述。
我熟悉供应商使用的方法，并不是最好的方法。
它肯定不到1：1的TUR。
他们所做的只是使用85054B校准套件标准化8510C，然后根据校准系数计算的预期值测量开放的反射相位。
我过去曾测量过很多这样的设备，虽然已经过了几年。
但是，如果存储器服务，反射相位在没有插入的情况下会有很大差异，因此它会在整个范围内发生故障，而不仅仅是1 GHz以上。
尽管如此，那只是关闭记忆而我可能错了。
我的初学者想法可能只是脏连接器或可能怀疑VNA电缆/适配器。
我对这种方法有多处保留，并一直在努力将所有这些试剂盒发送给安捷伦，但这是另一次讨论。
现在眼前的问题是如何量化失败的影响（假设它们实际上是合法的坏）。
在他们决定将其发送给安捷伦之前，拥有该套件的实验室只能假设该设备是坏的，并根据我们的质量保证政策采取下一步措施。
我将通过电子邮件发送拥有该工具包的实验室，看看我是否能获得数据，但我不希望在假期之后有任何回复。
如果我确实得到它，我可以坚持下去。
我有DrJoel的书，但在我完全消化和理解VNA测量背后的深层基本概念之前，我还有很长的路要走。
:)



     以下为原文

  Hey thanks for the response drkirkby.

So in actuality I simplified the scenario above.  What's actually happening is that this kit is in another one of our labs in a different location and they emailed me the question, so I don't have the data available, just the description above.  I'm familiar with the method used by the vendor, and its not the greatest by any means.  Its less than 1:1 TUR for sure.  All they do is standardize an 8510C with an 85054B cal kit, then measure the reflected phase of the open against the expected values calculated from the cal coefficients.

I've measured quite a few of these devices in the past, though its been a couple years.  However if memory serves, the reflected phase is quite a bit different without that insert, so it would have failed across the entire range, not just above 1 GHz.  Again though, that's just off memory and I could be wrong there.  My inital thought was probably just dirty connectors or maybe suspect VNA cables/adapters.

I have multiple reservations about the method and have been pushing to just have all these kits sent to Agilent, but that's a discussion for another time.  Right now the issue at hand is how to quantify the impact the failures have (assuming that they are in fact legitimately bad).  Until they decide to send them to Agilent, the lab that owns the kit can only assume the device is bad and take the next step per our QA policies.  I will email the lab that owns the kit and see if I can get data, but I wouldn't expect any response till after the holidays.  If I do get it, I can stick it up here.

I have DrJoel's book, but I've got a ways to go before I can fully digest and comprehend the deep underlying concepts behind VNA measurements.  :)

一些可能的情况是：它们没有正确插入扩展器（可能来自脏连接器，它们的公头类型 - N引脚与标准安捷伦引脚有一些区别，具体取决于它们使用的附件。或者：
他们用于初始校准的calkit未正确使用。或者他们没有使用8753定义的calkit系数（与8510或8720不同，即使是相同的部分）。



     以下为原文

  A few likely scenarios are : they left didn't properly insert the extender (can be from a dirty connector, of their male type-N pin is has some difference from the standard Agilent pin, depending on the adpater they use.

Or: the calkit they used for the initial calibration was not properly used.

Or they didn't use the 8753 defintiions for the calkit coefficients (which are different from the 8510 or 8720, even for the same part).

> {quote：title = Dr_joel写道：} {quote}>或者他们没有使用8753定义的calkit系数（与8510或8720不同，即使是同一部分）。
从哪里获得不同的定义？
我的8720D具有最新的固件，固件中有85032F，但*不​​是85032B。
所以我从校准套件手册中获取了85032B的系数。
我没有看到任何其他数据来源 - 可能的例外是过时的85054A校准套件的开放标准，我曾经使用过。
（哦，我希望我能负担得起85054B校准套件。圣诞老人并没有带给我一个圣诞节）。
DaveEdited：drkirkby于2013年12月25日下午12:03



     以下为原文

  > {quote:title=Dr_joel wrote:}{quote}

> Or they didn't use the 8753 defintiions for the calkit coefficients (which are different from the 8510 or 8720, even for the same part).

Where does one get the different definitions from? My 8720D, which has the latest firmware, has the 85032F in the firmware, but *not* the 85032B. So I tooks the 85032B's coefficients from the manual on the cal kit. I don't see any other source of data - with the possible exception of the open standards from an obsolete 85054A cal kit, which I have used on occasions. (Oh, I wish I could afford an 85054B cal kit. Santa did not bring me one for Christmas). 

Dave

Edited by: drkirkby on Dec 25, 2013 12:03 PM

嗨Dave，如果你从[http://na.tm.agilent.com/pna/caldefs/PNA/85032BE.htm]抓住了85032B系数，那么我相信那些将是你的8720D中输入的正确系数
那个工具包。
我认为Joel只是说，如果有一个8753，人们更愿意使用该套件的8753特定定义，因为那些针对操作进行了优化顺便说一下，在我上面引用的页面中，电子表格略微舍入了
男性打开 - 精确值（在该工具包的PNA文件中找到）是17.411 pSec（11被取消）。
此处显示的所有其他值与该工具包的PNA文件中显示的内容完全匹配。
节日快乐，布拉德



     以下为原文

  Hi Dave,

If you grabbed the 85032B coeffs from [http://na.tm.agilent.com/pna/caldefs/PNA/85032BE.htm], then I do believe those would be the correct coefficients to enter in your 8720D for that kit.  I think Joel's just saying that if one has an 8753, one would prefer to use the 8753-specific definitions for that kit since those were optimized for operation <= 6 GHz.

By the way, in the page I referenced above, the spreadsheet slightly rounded the Delay of the male Open -- the precise value (as is found in the PNA files for that kit) is 17.411 pSec (the 11 got dropped off).  All other values shown there match precisely with what are seen in the PNA file for that kit.

Happy holidays,
Brad

> {quote：title = bhokkan写道：} {quote}>嗨Dave，>>如果你从[http://na.tm.agilent.com/pna/caldefs/PNA/85032BE.htm]中抓取85032B系数，
那么我相信那些将是你的8720D中输入该套件的正确系数。
我没有从该文件中获取它们，而是从我附加的PDF手册中获取它们。
它们与您引用的文件中的相同，除了您提到的舍入错误。
>我认为Joel只是说，如果有一个8753，人们会更喜欢使用8753特定的那个套件定义，因为那些优化了操作节日快乐，> Brad你也是Brad.Edited：drkirkby于2013年12月27日
晚上11:53



     以下为原文

  > {quote:title=bhokkan wrote:}{quote}
> Hi Dave,
> 
> If you grabbed the 85032B coeffs from [http://na.tm.agilent.com/pna/caldefs/PNA/85032BE.htm], then I do believe those would be the correct coefficients to enter in your 8720D for that kit. 

I did not get them from that file, but from the PDF manual which I have attached. They are the same as in the file you reference, except for the rounding error you mention. 

>  I think Joel's just saying that if one has an 8753, one would prefer to use the 8753-specific definitions for that kit since those were optimized for operation Happy holidays,
> Brad

You too Brad.

Edited by: drkirkby on Dec 27, 2013 11:53 PM

附件

• 85032BE_type-N_6GHz_calibration-kit.pdf1,005.9 KB