Overlap Precedence如何在Momentum 2009中发挥作用？

重叠先行是相关的，其中来自不同绘图层的多边形映射到Momentum叠层中的相同z位置。
在您的情况下，顶部和底部导体必须位于叠层中的不同z位置，以正确表示电介质顶部和底部的金属。
你能告诉我们你的动量基底定义，即顶部和底部绘图层的映射位置吗？



     以下为原文

  Overlap preceeding is relevant where polygons from different drawing layers are mapped to the same z-position in the Momentum stackup.

In your case, top and bottom conductor must be at different z-positions in the stackup, to properly represent the metal at top and bottom of the dielectric. Can you show us your Momentum substrate defintion, i.e. where your top and bottom drawing layers are mapped?

Momentum是一个平面模拟器，其中电介质是无限的，因此您无法在迹线下方模拟有限的电介质宽度。
但我希望无限条带具有较低的共振频率，而不是更高，所以也许其他错误。
根据我在EM建模方面的经验，问题通常出在模型中而不是模拟工具中。
我会查看您的文件，但我的ADS2011 / 2012许可证不支持ADS2009，我认为* .dsn不足以进行故障排除，因为您还需要检查所有Momentum设置。
如果您使用Momentum设置归档项目，并上传* .zap，我可以将您的项目导入ADS2012并查看。



     以下为原文

  Momentum is a planar simulator where the dielectrics are infinite, so you cannot model the finite width of dielectric underneath the trace. But I would expect the infinite strip to have a lower resonance frequency, not higher, so maybe something else is wrong. From my experience in EM modelling, the problem is usually in the model and not in the simulation tool.

I would look at your file, but my ADS2011/2012 license does not support ADS2009 and I think that the *.dsn is not enough for troubleshooting because all your Momentum settings need to be checked as well. If you archive the project with your Momentum setup, and upload that *.zap, I can import your project into ADS2012 and have a look.

亲爱的沃尔克先生，非常感谢你跟进我的要求。
附件是我试图模拟的天线的Ring_Antenna.zap文件。
我使用ADS 2009归档了该项目，因此希望您可以将其导入到更新版本的ADS中。
我同意你的共振频率应该低一些不高。
就像一条信息一样，我使用Allegro设计天线几何，并将其导出为DXF文件。
然后我将它导入到ADS中，我想知道导入过程是否会导致它被误读或误解。
一旦您尝试将其加载到最新版本的ADS中，我一定会从您那里学到一些东西。
运行后，我是否可以将导出的项目从较高版本加载到较低版本？
无论如何我会尝试加载它，但也许你可以告诉我你在项目中看错了什么。
最好的问候，谢谢你的帮助rjrodrig



     以下为原文

  Dear Mr. Volker,

Thank you so much for following up on my request.  Attached is the Ring_Antenna.zap file for the antenna I am trying to simulate.  I archived the project using ADS 2009 so hopefully you can import it into a more recent version of ADS.  I do agree with you the resonant frequency should be lower not higher.  Just as a piece of information, I used Allegro to design the antenna geometry, and exported it as a DXF file.  I then imported it into ADS, and I wonder if the import process caused it to be misread or misinterpreted in anyway.

I will certainly learn something from you once you attempt to load it into your most recent version of ADS.  After you run it, will I be able to load your exported project from a higher version to a lower version?  I will try to load it anyway, but maybe you might be able to tell me what you see wrong on the project.

Best regards, and thank you for your help

rjrodrig

附件

• Ring_Antenna.zap853.5 KB
  

你好，我得到的结果与你的相似：低频电感相似，所以尺寸似乎没问题，但SRF是不同的。
不知道那里发生了什么。
EM模型和设置对我来说很好。
最好的问候Volker



     以下为原文

  Hello again,

the results that I get are similar to yours: Low frequency inductance is similar, so the dimensions seem to be ok, but the SRF is different.

Not sure what is going on there. The EM model and setup looks fine to me. 

Best regards
Volker

附件

• loop_results.PNG35.4 KB
  

非常感谢您模拟和确认我的结果。
至少我没有做错什么。
您在顶层和底层使用了哪些优先数字？
有关系吗？
通常它默认为0，但无论数字如何，它都没有模拟我正在做的模拟。
我非常感谢您使用更新的工具模拟此结构，并确认我的ADS2009答案。
在任何情况下，我认为是时候切换到更准确的工具，可以预测接近现实生活的平面结构共振，就像CST EM那样。
它看起来像CST EM是要走的路，我将不再浪费我的时间ADS进行这种分析。
它离开了，我永远不会自信地设计天线，因为我的测量值永远不会与模拟结果一致。
老实说我对这个工具很失望，我希望是我做错了。
我不确定CST是如何做到的，但他们每次都给出了正确答案。
他们的结果与我用VNA收集的测量数据相匹配。
ADS到处都是......好吧Ce la vie！
非常感谢你的帮助！
编辑：rjrodrig于2013年2月21日下午10:27



     以下为原文

  Thank you so much for simulating and confirming my results.  At least I am not doing something wrong.  What precedence numbers did you use for top layer and bottom layer?  Does it matter?  Normally it defaults to 0, but it has no baring on the simulation I am doing regardless of the number.  I do appreciate your time simulating this structure with a more up-to-date tool, and confirm my ADS2009 answer.

In any case, I believe it is time to switch to a more accurate tool that can predict planar structure resonances close to real life like CST EM does it.  It looks like CST EM is the way to go and I will stop wasting my time ADS for this kind of analysis.  It is way off,  and I will never design an antenna with confidence as my measured values are never going to agree with the simulated results.  

I am honestly disappointed in the tool, and I was hoping it was me doing something wrong.  I am not sure how CST does it, but they nail the correct answer every time.  Their results match the measured data I gathered with the VNA.  ADS is all over the place...

Oh well Ce la vie!

Thanks a bunch for your help!

Edited by: rjrodrig on Feb 21, 2013 10:27 PM

正如我在第一篇文章中提到的，优先级是指来自不同层的重叠对象，这些对象被映射到相同的z位置。
这不适用于您的模型。
我知道您对测量结果缺乏一致感到失望，但我不明白为什么Momentum会在这里给出不正确的结果，除非您已经模拟了与测量结果不同的东西。
为了仔细检查Momentum结果，我在Sonnet EM模拟器中运行了相同的几何体，这是非常准确的。
通过Sonnet，我得到了与Momentum非常相似的结果。
所以无论问题是什么，我都怀疑它是Momentum的准确性。
您是否在CST中对此精确几何进行了建模，或者CST模型是否包含一些额外的细节（除了有限的基板形状）？
您能否在CST中显示CST模型和端口详细信息的图片？



     以下为原文

  As mentioned in my first post, precedence refers to overlapping objects from different layers that are mapped to the same z-position. This does not apply to your model. 

I understand that you are disappointed by the lack of agreement to measurement results, but I can't see why Momentum would give incorrect results here, unless you have modelled something different from what you have measured. 

To double check the Momentum result, I ran the same geometry in the Sonnet EM simulator, which is know to be very accurate. With Sonnet, I get get very similar results to Momentum. So whatever the problem is, I doubt that it's Momentum's accuracy.

Did you model this exact geometry in CST, or did the CST model include some additional detail (in addition to the finite substrate shape)? Can you show a picture of the CST model and the port details in CST?

相当于CST MWS的安捷伦相当于EMPro，而不是ADS。
使用ADS，您可以使用Momentum模拟具有无限基板尺寸的PCB，或者使用FEM求解器（需要许可证）假设有限的基板尺寸。
然而，我已经在EM建模和EM支持方面工作了14年，我真的怀疑这种差异与有限的基板尺寸有关，并且怀疑CST模型包含ADS布局中缺少的其他一些提升细节。
要找到答案，您应该联系您的CST许可证，并要求他建模您的ADS布局（不是更多，而不是更少），有/无有限基板。
这样，您可以将苹果与苹果进行比较，并了解差异的来源。



     以下为原文

  The Agilent equivalent to CST MWS would be EMPro, not ADS. 

With ADS, you can simulate the PCB assuming infinite substrate size with Momentum, or assuming finite substrate size with the FEM solver (license required).

However, having worked 14 years in EM modelling and EM support, I really doubt that the difference is related to finite substrate size, and suspect that the CST model includes some other elevant detail that is missing in your ADS layout. To find out, you should contact your fried with the CST license, and ask him to model your ADS layout (not more, not less) with/without finite substrate. That way, you can compare apples to apples, and know where the difference comes from.

除了我以前的帖子：我现在用ADS FEM模拟你的天线，有和没有有限的PCB形状。
凭借无限的PCB，SRF与Momentum完美契合。
正如预期的那样，当使用切口模拟PCB而不是全（无限）PCB时，自谐振频率上升几MHz。
然而，效果很小。
现在我们有3种不同的EM模拟器（ADS Momentum，ADS FEM，Sonnet），它们都为您的布局提供> 70MHz的SRF结果。
与您的朋友CST模拟的差异必须在几何中，而不是由模拟引擎引起。
有些东西不同或缺失。编辑：volker_muehlhaus于2013年2月24日上午7:44



     以下为原文

  addition to my previous post:

I now simulated your antenna in ADS FEM, with and without the finite PCB shape. With infinite PCB, the SRF agrees well with Momentum. As expected, the self resonance frequency goes UP a few MHz when the PCB is simulated with cutout, instead of full (infinite) PCB. This is effect is small, however. 

Now we have 3 different EM simulators (ADS Momentum, ADS FEM, Sonnet) which all give SRF results > 70MHz for your layout. The difference to your friends CST simulations must be in the geometry, and not caused by the simulation engine. Something is different or missing.

Edited by: volker_muehlhaus on Feb 24, 2013 7:44 AM

亲爱的沃尔克先生，我一直在说必须有一个设置差异或其他东西。
我拜访了我在奥地利的朋友，他提到没有自己模拟天线。
他为它添加了同轴电缆模型。
这就是他的模拟与VNA测量一致的原因。
所以，你是对的。
他增加了一个传输线模型，将谐振频率降至48MHz。
这是你和我一直追逐的谜。
因此，我现在相信Momentum的输出，我敢打赌，如果我添加同轴模型，我会在动量上测量相同的。
在一个完全独立的说明中：动量可以让我将两个天线结构相互叠加在一起吗？
如何从左到右，顶部和底部滑动顶部天线以参数化地模拟两个结构之间的耦合？
这是可能的Momentum还是我需要其他工具？
如果Momentum可以做到，你能分享一下这样做吗？
在这种情况下，我非常感谢你的所有工作。



     以下为原文

  Dear Mr. Volker,

I have been saying that there must be a setting difference or something.  I visited my friend in Austria, and he mentioned the didn't model the antenna on its own.  He added a coaxial cable model to it.  That is why his simulation agrees with the VNA measurement.  So, you are correct.  He added a transmission line model which brought down the resonant frequency to 48MHz.  

This was the mystery you and I have been chasing.  As such, I now believe Momentum's output and I bet if I add the coaxial model I will measure the same in momentum.

On a completely separate note:  Can momentum allow me to model two antenna structures one on top each other?  How can I slide the top antenna left to right and top and bottom to model the coupling between the two structures parametrically?  Is this possible in Momentum or do I need another tool?  If Momentum can do it, can you share how to do so?

I really appreciate all your work in this case.