为什么VNA校准和验证套件价格如此之高

> {quote：title = theamberco写道：} {quote}>嗨，出于好奇我不知道是否有人会费心去回答这个问题，但是第一次看到校准套件时，我的价格让我感到震惊。
实际上从外面的材料看起来像钢连接器，泡沫，一个漂亮的木盒子。
我只能想象，对于重复使用的公差，材料的等级有多精确，可能是高价来自或可能是因为它的独特性。
可能他们是由我不知道的专家手工制作的。
制造这种试剂盒需要什么样的特殊设备，内部是否有任何超昂贵的材料，安捷伦是否将它们作为第三方使用。
基础材料是铍铜（非钢），具有特殊的多层镀金。
我们在特殊的瑞士机器上制造我们称之为“钻石铣削”中心的大部分零件。
有一段时间我们考虑将机器搬到一个新的（低成本）设施，但是机器需要花费几个月的时间才能达到制造零件所需的精度。
我认为转向规格是20微英寸（小于1微米）。
关键规格是将中心导体转换为精确的直径，我们认为我们可以使它们比世界上任何其他地方都更好（他们总是在Santa Rosa和瑞士国家标准组织之间进行一些竞争，关于谁可以制造一个
最准确地测量它们）。
事实上，该过程的很大一部分是开发测量能力。
一旦零件转动，你必须采用精密电镀（我们有自己的电镀车间，在加利福尼亚州，由于环境问题，这是非常昂贵的），以实现正确的最终尺寸。
并且电镀具有特定的硬度以允许重复连接。
然后，连接器非常复杂;
它是校准套件上的无槽连接器，由安捷伦再次开发（当时惠普，由我的导师Julius Botka和VNA桥梁设计师开发，直到我为他接手---
好吧，我和另一个乔尔;有一个笑话，你必须被命名为乔尔，在桥梁和耦合器上工作一段时间）。
> VNA是一种先进的电子机器，内部必须具有精密机械部件，并且设计复杂，我觉得它们的价格不如cal套件高。
事实上，校准套件中的部件比VNA中的部件精确至少10或100倍...... VNA几乎可以从校准套件中获得所有精度。
>另外我知道有些机器内部有工厂校准，如果没有标准可以召回，粗测量会有更多的不确定性。
但是有可能在电缆末端使VNA完全自校准，也许可以使用VNA中自包含的标准。
与现在的校准和验证套件和程序相比，这是否更实惠，更简单？
我们刚刚介绍了这样的产品，但它位于测试电缆的末端，因为测试电缆是任何测试系统中最糟糕的部分;
请参阅http://na.tm.agilent.com/pna/calpod/>顺便说一句，我喜欢木盒子，希望安捷伦不要用塑料盒来改变它们。
他们会留下来。
但奇怪的是，在Ecal（电子校准）上，它们是一种选择。
长篇故事也是......



     以下为原文

  > {quote:title=theamberco wrote:}{quote}
> Hi, just out of curiosity I dont know if anyone would bother to answer this but the first time I looked at a calibration kit and the price at I was shocked. Actually from the outside the materials were a few look like steel connectors, foam, a nice wooden box. I can only imagine how precise has to be the tolerances, grade of materials for repeated use and probably that is where the high price comes from or maybe is because of its uniqueness. Probably they are hand made by experts I just dont know. What kind of special equipment is necessary to make this kits, is there any ultra expensive materials inside and does Agilent makes them o some third party. 
The base material is beryllium-copper (not steel) with a special propritery multilayer gold plating.  We make most of the parts in what we call our "diamond milling" center, on special Swiss machines.  At one time we looked at moving the machines to a new (lower cost) facility but it would take something like a few months for the machines to settle down enough to achieve the precision needed to make the parts.  I think the turning spec is something like 20 micro-inches (less than 1 micron) .  The key spec is turning the center conductors to a precise diameter, and we think we can make them better than any place else in the world (their's always some competition between us in Santa Rosa, and the Swiss national standards organization about who can make a measure them most precisely).  In fact, a big part of the process is developing the measurment capabilty.  And once the part is turned, you have to apply a precision plating (we have our own plating shop, which in california is ridiculously expensive due to environmental concerns) to acheive the proper final dimensions.  And the plating has to a specific hardness to allow repeated connections.
Then, the connector is very complicated; it is a slotless connector on the cal-kits, again developed by Agilent (well, HP at the time, by Julius Botka, one of my mentors, and the designer of the bridges in VNA, until I took over for him --- well me and another Joel; there was a joke that you had to be named Joel to work on bridges and couplers for a while).
> A VNA is a state of the art electronics machine that must have precision mechanical parts inside as well and with its design complexity I feel they are not as highly priced as cal kits. 
In fact the parts in the cal kit are at least 10 or 100 times more precise than the parts in the VNA...the VNA gets almost all its accuracy from the cal kit.
> Also I know some machines have a factory cal inside that can be recalled in case they is no standards available, of coarse measurements will have more uncertainties. But would it be possible to make a VNA totally self calibrating at the end of the cables maybe with the standards self contained in the VNA. Could this be more affordable and less cumbersome than it is now with the calibration and verification kits and procedures?
We just introduced such a product but it goes at the ends of the test cables, since test cables are the worst part of any test system; see http://na.tm.agilent.com/pna/calpod/
> By the way I do love the wooden boxes I hope Agilent dont change them with plastic ones.
They will remain.  But oddly, on the Ecal (electronic calibration) they are an option. Long story for that too...