为什么VNA不需要3.5毫米校准标准的性别

Dr_joel，我不同意。
我认为一些想法非常聪明，而且并不明显。
核磁共振（或现在称为MRI），非常聪明，完全不明显。
（也许这就是发明者获得诺贝尔奖的原因）。
我不太相信男性和女性有相同的延迟，所以没有必要指定性别..（我知道我冒着你告诉我你发明了这种方法的风险）。
谢谢你的论文 - 我会读一读，毫无疑问会有一些问题。
我有兴趣了解使用完整的3D EM仿真器（如Agilent的EMPro，Ansys的HFSS或类似设备）来获取系数的可能性。
我不知道是否有人曾经这样做过，但即使不是，我也不会认为这是一个特别聪明的主意。
它是否有效是另一回事。
如果确实如此，它可以让我找到我的pameters我订购了你的书的副本 - 毫无疑问我会有一些问题。
我在那里收集了关于TDR操作的相当多的内容，我会觉得有趣，因为我的8720D确实有TDR选项（010），但我还没有使用它。
戴夫



     以下为原文

  Dr_joel, 
I disagree. Some ideas I think are very clever, and not obvious. NMR (or MRI as it is called now), is very clever, and totally non-obvious. (Perhaps that is why the inventor got a nobel prize for it). 

I'm less convinced about having the male and female having the same delay, so there's no need to specify the sex.. (I know I risk you telling me you invented this method). 

Thank you for the paper - I will take a read, and no doubt have some questions.

I'd be intersted to find out how close it is possibe to get the coefficients using a full 3D EM simulator like Agilent's EMPro, Ansys's HFSS or similar. I don't know if anyone has ever done that, but even if not, I would not consider it a particulary clever idea. Whether it would work is another matter. If it did, it may allow me to find pameters for my 

I have ordered a copy of your book - no doubt again I will have some questions. I gather there is quite a bit in there about the TDR operation, which I'll find interesting as my 8720D does have the TDR option (010), but I have not used it yet.  

Dave

（事实上​​，我大部分时间都在嘲笑男性和女性标准与同一模型相匹配的聪明才智，不，我不负责任）。
至于使用3D模拟，是的，这样做是实际和合理的。
事实上，我负责开发75欧姆连接器母头开放型号，用于新的4槽长指连接器（如果误插入50欧姆N型连接器，它可以存活）。
为了设置边界条件，我设置了两个模型：一个用于实现最大相移，这意味着最大电容，因此这是最小直径外导体，最大直径中心导体，手指最大扩展，手指最宽，最小插槽宽度。
所有这些都取自底层部件的物理尺寸限制。
对于最小相移，我选择了相反的尺寸，最短的线，最宽的槽等。这给出了相位与频率的范围。
我选择中点进行曲线拟合以获得多项式的Cn值，然后我计算了与中点的偏差，以计算设置源匹配规范的最大误差。
几个月后，生产航空公司以验证结果，偏差中点提供了对实际性能的良好估计。
这种技术最坏的情况;
一个人可以合理地使用最大/分钟的标准偏差估计进行相同的过程，以获得更具统计性的驱动值;
如果原始值被认为过大，则使用95％覆盖因子（K = 2.77）可能会产生较小的源匹配估计值。



     以下为原文

  (In truth, I was mostly jesting about the cleverness of making male and female standards match the same model, and no, I wasn't responsible for that).

As to using 3D simulation, yes it is practical and reasonable to do this.  In fact, I was responsible for developing the 75 ohm connector female open model, for a new 4-slot long-finger connector (it could survive if one mistakenly inserted a 50 ohm type N connector).  To set the boundary conditions I setup two models:

One to acheive maximum phase shift, which means maximum capacitance, so this was smallest diamter outter conductor, largest diameter center conductor, maximum spread of the fingers, logest fingers, minimum slot width. All this was taken from the physical dimension limits of the underlying parts.  For minimum phase shift, I chose the opposite dimensions, shortest lines, widest slots, etc.

This gave a range of phase vs frequency.  I chose the mid point to do a curve fit to obtain the Cn values for the polynomial then I computed the deviation from the mid point to compute the maximum error from which the specifications for source match were set.  Several months later, airlines were produced to verify the results and the deviation midpoint provided a good estimate of the actual performance.

This technique gives worst case; one might reasonably do the same process using an estimate of standard deviation insteate of max/min to obtain a more statiistically driven value; using a 95 percent coverage factor (K=2.77) might yield a smaller estimate for source match in that case if the original value was deemed execessively large.