怎么确保功率限制器正常工作

嗨 - 正如您所发现的，11867A RF限制器在此应用中使用时已损坏。
不幸的是，限制器被损坏了，但它确实保护了信号分析仪的前端，这比修复限制器的成本要多得多。
我意识到11867A限幅器看起来像一个非常昂贵的RF保险丝（安捷伦约500.00美元）。
更换分析仪前端包括调整和验证非常昂贵。
当DUT的电源状态发生变化时，LISN器件会产生巨大的尖峰。
11867A射频限制器（平均每瓦10瓦）不像其他设备那样强大，例如Agilent 11947A瞬态限制器（10千瓦10安）。
但是，Agilent 11947A的频率范围为9 kHz至200 MHz，对于50 Hz应用来说，这还不够低。
我做了一些研究，发现安捷伦不再生产11947A瞬态限制器，但是如果你去安捷伦11947A瞬态限制网页上有一个链接到Electro-Metrics，他们的型号EM7600瞬态限制器似乎有相同的
作为11947A，10 us的10 kW值。
请查看此瞬态限制器，看看是否可以使用更高频率的测试。
我记得几年前，当我看到我们的环境实验室使用LISN测试EMI / EMC时，为了更加安全，控制测试的软件指示技术人员断开频谱分析仪的输入，然后更改DUT设置，然后重新连接
分析仪输入，这样做是为了更加安全。
以下是11867A限制器的说明：限制器是单向的（标有输入端口和输出端口），因为信号看到的第一组二极管是PIN，因为它们可以处理比Schottkys更多的功率。
肖特基为PIN生成的DC提供DC返回路径。
Schottkys最初是为了获得最初的尖峰并开始产生直流电流，然后是PIN。
因此，限制器中反映的大部分功率在PIN中消散。
请参阅11867A操作和服务指南：http：//www.home.agilent.com/agilent/facet.jspx？＆amp; cc = US＆amp; lc = eng＆amp; amp; k = 11867-90005＆amp; sm = g可能发生的事情
你的限制器是一个或多个二极管损坏而且没有完全失效。
这将导致5 dB的插入损耗。
11867A不可修复，需要更换。
确保插入损耗正确的唯一方法是使用可测量插入损耗的测试系统。
确保输入和输出端口提供50欧姆匹配。
入射功率必须低于1mW限制阈值。
问候 -



     以下为原文

  Hi -

As you discovered, the 11867A RF limiter was damaged when used in this application. It is unfortunate the limiter was damaged, but it did protect the front end of the signal analyzer, which would cost much more to repair than the cost of replacing the limiter. I realize the 11867A limiter seems like  a very expensive RF fuse (~500.00 US from Agilent). Replacing the analyzer front end including adjustment and verification is really expensive.

LISN devices can put out huge spikes when the power state of the DUT changes. The 11867A RF limiter (10 W avg.) is not as  robust as other devices such as the Agilent 11947A Transient Limiter (10 kW for 10 us).  However the Agilent 11947A has a frequency range from 9 kHz to 200 MHz, which is not quite low enough for your 50 Hz application. 
I did some research and found the 11947A Transient Limiter is no longer being produced at Agilent, but if you go the Agilent 11947A Transient Limit web page there is a link to Electro-Metrics for their model number EM7600 Transient Limiter that appears to have the same 10 kW for 10 us values as the  11947A. Please take a look at this transient limiter to see if it can be of use if you have tests at higher frequencies.  

I recall years ago when I watched our Environmental lab test EMI/EMC using a LISN, just to be extra safe, the software controlling the test instructed the technician to disconnect the input to the spectrum analyzer and then change the DUT settings, then reconnect the analyzer input, and this was done to be extra safe.  

Here is the description of the 11867A limiter:
The limiter is unidirectional (has input port and output ports marked) because the first set of diodes the signal sees are the PINs since they can handle more power than the Schottkys.
Schottkys provide DC return path for the DC generated by the PINs. The Schottkys come on initially to get the initial spike and start generating DC current, then the PINs come on.
So, the majority of the power that is reflected in the limiter is dissipated in the PINs.

Please see the 11867A Operating and Service Guide: 

http://www.home.agilent.com/agilent/facet.jspx?&cc=US&lc=eng&k=11867-90005&sm=g

What probably happened to your limiter is that one or more of the diodes was damaged and did not completely fail. This would account for the 5 dB insertion loss. The 11867A is not repairable, it needs to be replaced.

The only way to assure the insertion loss is correct is to use a test system that can measure insertion loss. Assure the input and output ports are provided with a 50 ohm match. Incident power must be below the 1mW limiting threshold.

Regards -

您好Tabbot，非常感谢您的有用帮助和其他信息。
我会看看你建议的更强大的限制器，即使我还要找到一个较低切割频率的匹配。
是的，我们对接收器前端之前的射频功率限制器感到高兴...最后我们保存了它。
您的解释也帮助我更好地理解如何只能看到几个dB衰减而不是类似于开路或闭路的条件。
我们已经尝试采取所有预防措施以避免损坏我们的仪器，包括从“噪声参考”测量开始（DUT未连接到LISN）并在更换被测lisn时断开接收器但在这种情况下他们没有
够了。
我想在不久的将来我们将开始在lins和rf功率限制器之间使用更高的衰减器。
如果您有任何其他建议进行安全测量，我将不胜感激。
最好的祝福



     以下为原文

  Hi Tabbot,

Thanks a lot for your useful help and additional information as well. I'll take a look to the more powerful limiters you have suggested, even if I've also to find a match with the lower cutting frequency.

Yes, we are happy about our RF power limiter before the front end of the receiver... at the end we have saved it.

Your explanation also helped me to better understand how it has been possible to see only few dBs attenuation instead of a condition similar to open or close circuit. We've tried to take all the precautions to avoid damages to our instrumentation, including starting with "noise reference" measurements (with DUT not connected to the LISNs) and disconnect the receiver when changing the lisn under test but in this case they have not been enough. I think in the close future we will start using also higher attenuators between lins and rf power limiter.

If you have any other suggestion to perfom safe measurement, I'll appreciate it.

Best Regards