HP 11850C电源分配器标记为“OUTPUT”的端口是否带有贴纸

> {quote：title = SOLT_guy写道：} {quote}>尊敬的先生：>> HP 11850C功率分配器被宣传为“三路功率”分配器。
>关于此设备的文档很少，我不想对设备做任何假设。
>任何人都可以告诉我，标有“OUTPUT”标记的端口是否与未标记的端口配置不同（此设备是否缺少标记输出端口的贴纸）？
或者，标记为“OUTPUT”的端口是否用于特殊目的（例如被指定连接到网络分析仪参考信号端口）？
我不知道该设备，我注意到你说你不想做出假设，但我认为以下是非常合理的假设，并且应该允许你确定设备的功能。
* 11850C工作在DC的事实意味着它必须是电阻功率分配器，而不是基于变压器的功率分配器。
* 11850C具有电阻性，这意味着您可以用万用表探测它，看看它的功能。
*对于设计为使得S11 = S22 = S33 = S44的3路电阻功率分配器，理论损耗为9.54dB。
http://www.microwaves101.com/encyclopedias/resistive-power-splitters鉴于该设备的丢失据说是9.5 dB + 1dB / GHz，这表明我的设计是S11 = S22 = S33 = S44。
有两种制作电阻式功率分配器的方法。
*怀伊安排。
*三角洲安排。
制造三路功率分配器最明显的方法是采用四个电阻器的星形/星形布置。
它们的值应该是R = Z0x（N-1）/（N + 1），因此对于3路功率分配器，它将在一颗恒星中具有四个25欧姆的电阻器。
如果它是这样构建的，那么用万用表测量任意两个中心导体，而在其他连接器上没有任何东西，你将测量50欧姆，因为它将有两个串联的25欧姆电阻器。
在内导体和外导体之间，它将显示开路。
三角形布置将使用万用表显示不同的行为。
我会留下你来计算电阻值！
这当然不是一件轻而易举的事。
我见过一些双路HP 26.5 GHz电阻式功率分配器，它们在所有端口都不匹配，但只有几个电阻器连接在输入端口。
我认为这会带来更低的损失。
但我不认为11850C就是这种情况，因为损耗值与设计的理论值相匹配，无论如何设计似乎是最明显的设计。
我想你应该能用万用表来解决这个问题。
3458A可以很好地完成这项工作，但我认为它只是有点矫枉过正。
我认为一个3.5位数的手持式万用表足以解决该设备的问题。
但最终你可以在VNA上测试它。
然后你就确切地知道它是如何表现的，这可能与某人有损坏它应该做的事情有所不同。
但主要基于是德科技网站上给出的损失数字，我认为除了对称安排之外，它不太可能是其他任何东西。
毫无疑问有些不对称，特别是在输入端口周围，因为它在另一个面上与其他人不同，但我看不出有多大的影响。
Dave编辑：drkirkby于2016年3月13日上午10:18编辑：drkirkby于2016年3月13日下午12:08编辑：drkirkby于2016年3月13日下午12:10



     以下为原文

  > {quote:title=SOLT_guy wrote:}{quote}
> Dear Sir:
> 
>       The HP 11850C power splitter is advertised as a "three way power" splitter.  
 

>        There is little documentation available about this device and I do not like to make any assumptions about a device.   

>         Can anyone tell me if the port labeled, "OUTPUT" with a sticker on it, is configured differently from the ports that are not labeled at all (is this device just missing the stickers that label the output ports)?   Or, is the port labeled, "OUTPUT", intended to be used for a unique purpose (such as being designated to be connected into a network analyzer reference signal port)?

I don't know the device, and I note you say you don't want to make assumptions, but I think the following are very reasonable assumptions, and should allow you to determine what the device does. 

* The fact the 11850C works down to DC means it must be resistive power splitter, rather than a transformer based power splitter.  
* The fact the 11850C is resistive means you can probe it with a multimeter to see what it does. 
* For a 3-way resistive power splitter designed such that S11=S22=S33=S44, the theoretical loss is 9.54 dB. 

http://www.microwaves101.com/encyclopedias/resistive-power-splitters

Given the loss of that device is said to be 9.5 dB + 1dB/GHz that suggests to me it is designed to have S11=S22=S33=S44. 

There are two ways of making resistive power splitters. 

* Wye arrangement. 
* Delta arrangement. 

The most obvious way to make a 3-way power splitter is a star/wye arrangement of four resistors. Their values should be R=Z0x(N-1)/(N+1), so for a 3-way power splitter it would have four 25 Ohm resistors in a star.  If it is built like that, then measuring between any two centre conductors with a multimeter, with nothing on the other connectors, you will measure 50 Ohms, as it will have two 25 Ohm resistors in series. Between inner and outer conductors it would show an open circuit. 

The delta arrangement would show a different behavior with a multimeter. I will leave you to work out the resistor values! That's certainly non-trivial to do. 

I have seen some 2-way HP 26.5 GHz resistive power splitters which are *not* matched at all ports, but have a couple of resistors joined at the input port only. That I assume gives lower loss. But I don't think that's the case with the 11850C because the loss value matches the theoretical value for a design that seems the most obvious design anyway. 

I think you should be able to resolve this with a multimeter. A 3458A would do the job nicely, but I think its just a little bit overkill. I think a 3.5 digit handheld multimeter is good enough to sort out what that device is. But ultimately you can test it on a VNA. Then you know exactly how it is behaving, which might well be different from what it is supposed to do if someone has damaged it. But based primarily on that loss figure given on the Keysight website, I would think it is highly unlikely it is intended to be anything other than a symmetrical arrangement. The is no doubt some asymmetry, especially around the input port as it on another face to the others, but I can't see that having much effect. 

Dave

Edited by: drkirkby on Mar 13, 2016 10:18 AM

Edited by: drkirkby on Mar 13, 2016 12:08 PM

Edited by: drkirkby on Mar 13, 2016 12:10 PM

亲爱的柯比博士;
哇，你是过去的声音！
这么多年后，很高兴收到你的来信。
我不得不把我的微波学术兴趣搁置很长很长一段时间，因为我必须学习其他东西。
那是在我获得HP 11850C“Power Splitter”之前。
当我读你的帖子时，哇，我不知所措。
你很擅长分析。
我只是查看了产品的文档。
意识到没有带有原理图的文档，然后我意识到我正在研究什么是“内部”问题。
如果我可以通过写入Keysight / HP / Agilent来避免理论化“盒子内部”并获得经过验证的事实答案，我将写信给他们。
当我在损坏之前读取设备的低功率限制（我认为它是1瓦特）时就达成了协议。
当您尝试使用昂贵的硬件而刚刚拔出包装然后您意识到您损坏了它，因为您假设设备是如何工作的并且这是错误的假设时，没有什么能像您那种下沉的感觉。
我认为您的分析可能是正确的，但我需要验证。
我不明白为什么没有为此设备提供原理图，Keysight为其所有其他“分离器”提供原理图。
顺便说一下，在射频行业中，术语“分离器”和“功率分配器”在某个时间或每个人之间可以互换使用。
这为那些试图自己学习这些设备的人们带来了真正的问题，因为互联网上的信息并没有以正式的数学方式表达，还有列举的例子。
这留下了很多不确定性和解释空间（正确和错误的解释）以及试图使术语“笔直”的人头痛，并将它们与正确的模型相匹配，这种模型往往是，而不是由
信息来源。
为了更好地理解我的意思，请访问以下链接：http：//literature.cdn.keysight.com/litweb/pdf/5989-6699EN.pdf？id = 1261867“功率分配器与功率分配器之间的应用差异”
是德科技的某个人认识到可以互换使用术语“功率分配器”和“功率分配器”的问题，并写了一篇关于这两个术语之间差异的文章。
由于您不是来自美国，因此您使用这些条款没有实际意义（不适用），因为您来自不同的国家/地区，他们的语言和教育体系略有不同。
换句话说，我认识到你使用美国历史上可以互换使用的术语进行了不同的教学。
无论如何，我猜我们都期待在下一篇文章中找到“内容”。



     以下为原文

  Dear Dr. Kirby;

     Wow, are you a voice from the past!   It is good to hear from you after so many years.  

     I had to put my microwave academic interests on hold for a long, long, time because I had to learn other things.  That was until I acquired the HP 11850C "Power Splitter."

     When I read your post, wow, was I overwhelmed.   You are excellent at analysis.    

     I just looked at the documentation on the product.  Realized that there was no documentation with a schematic diagram and then I realized I am looking at a what is "inside the box" problem.  

     If I can avoid theorizing about what is "inside the box" and get a verified, factual, answer just by writing to Keysight/HP/Agilent, I am going to write to them.  When I read the low power limit of the device before damage (I believe it was 1 watt) that sealed the deal.  There is nothing like that sinking feeling you get when you experiment with an expensive piece of hardware that you just pulled out the packaging and then you realize that you damaged it because you made an assumption about how the device works and it was the wrong assumtption.   

     I think that your analysis may be correct but I need verification.   I don't understand why there is no schematic provided for this device, Keysight provides schematics for all its other "splitters".

     By the way, in the RF industry, the term power "splitter" and "power divider" are, or were, used interchangeably, by everyone at one time or another.   This created a real problem for people trying to learn on their own about these devices because the information on the internet was not expressed in a formal, mathematical manner, along with enumerated examples.   This left a lot of uncertainty and room for interpretation (correct and incorrect interpretations) along with a raging headache for the person trying to get terms "straight" and match them to the proper model which was more often, than not, not supplied by the information source.   

     To get a better understanding of what I mean go to this link:

http://literature.cdn.keysight.com/litweb/pdf/5989-6699EN.pdf?id=1261867

"Differences in Application Between Power Dividers and Power Splitter"

Someone at Keysight recognized the problem of using the terms "power divider" and "power splitter" interchangeably and wrote an article about the differences between those two terms.  

     Since you are not from the US, your use of those terms, is moot (not applicable), because you are from a different country where they have a slightly different language and educational system.  In other words, I recognize that you were taught differently using terms that were historically used in the US interchangeably.  

     Anyway, I gues we are both looking forward to finding out "what" is inside the box in the next post.

我记得在70年代中后期看到那个设计这个的家伙。
那时，它仅达到1.3 GHz（对于8505A。）他必须将四个1/8瓦的电阻器焊接在一起并将它们插入该主体的中心。
如果你想进行测量，你可以找到（数字舍入）：输入到任何其他端口：67欧姆输出端口到任何其他输出端口：100欧姆输入端口接地，所有3个输出端口都有短路：33
欧姆输入端口接地，所有3个输出端口均为50欧姆负载：50欧姆（正如预期保持良好匹配）唯一可行的配置是与输入端口串联的17欧姆电阻器和三个50欧姆电阻器，每个都在
每个输出端口的系列;
所有人都在中心联系在一起。
您的拆分器必须缺少两个“输出”标签;
我有它们。



     以下为原文

  I remember watching the guy that designed this back in the mid-late 70's. At that time, it only went to 1.3 GHz (for the 8505A.)  He had to solder four tiny 1/8 watt resistors together and insert them into the center of that body. 

If you want to do the measurements, this is what you will find (numbers rounded):

Input to any other port: 67 Ohms
Output port to any other output port: 100 Ohms
Input port to ground with shorts on all 3 output ports: 33 Ohms
Input port to ground with 50 Ohm loads on all 3 output ports: 50 Ohms  (as expected to maintain good match)

The only configuration that will work is a 17 Ohm resistor in series with the input port and three 50 Ohm resistors, each in series with each of the output ports; all tied together at the center.
Your splitter must be missing two of the "Output" labels; mine has them.

Jvall;
你的解释是如此“在金钱上”。
这正是我所测量的。
我想这个谜就解决了。
根据我刚刚提到的Keysight文章，我被认为HP11850C今天被归类为“功率分配器” - 而不是“功率分配器”。
哇，有时候变了！
Jvall，当你提到在70年代看到那个人在HP11850C上工作时 - 我所能做的只是笑。
七十年代是如此的爆炸！
谢谢你，Jvall，你的出色解释 - 它确实是 - “对钱”！



     以下为原文

  Jvall;

      Your explanation is so "on the money."  That is exactly what I measured.   

       I guess this mystery is solved.     I am led to believe the HP11850C is classified today as "power divider" - not a "power splitter" - according to the Keysight article that I had just mentioned.   Wow, have times have changed!

       Jvall, when you mentioned seeing the guy work on HP11850C in the 70's - all I can do is laugh.   The seventies was such a blast!!!

       Thank you, Jvall, for your excellent explanation - it really was -- "right on the money" !!!

是的，一个真正的功率分配器需要一个虚拟接地，这是没有的，所以它应该被称为功率分配器。
现在，如果我能及时回到大约40年，我会建议。
至少我不再有白头发了。



     以下为原文

  Yes, a true power splitter requires a virtual ground which this does not have, so it should be called a power divider.  Now if I can just go back in time about 40 years, I will suggest that.  At least I would no longer have grey hair.