PNA X具有反向扫描的增强响应ECAL问题

PNA-X是N5242A FW是9.42.16对于校准：我要求PNA-X管理校准（即我不执行2端口校准，删除反向错误术语并试图欺骗分析仪进行思考
它是一个增强的响应cal）。
如果我从前面板开始，我会选择Cal Wizard-> Ecal->检查src / rx power cal->从选择中选择'Defined Thru'和Enhanced Resp cal然后关闭我去....
我将尝试解耦端口电源并将端口2电源设置为合法的衰减值，如果这让我解决了问题，一切都很好。
然而，理解发生了什么事情会很好，因为虽然EnhResp cal在校准期间不需要从Port2驱动....'cal all'听起来像一个很棒的功能 - 我记得评论过一个安捷伦代表
2年前，这对于多通道配置来说是一个很好的功能，因为通常有很多冗余的重新连接...... Thx！



     以下为原文

  PNA-X is a N5242A
FW is 9.42.16

For the Calibration:
I am asking the PNA-X to manage the calibration (i.e. I am NOT performing a 2 port cal, dropping the reverse error terms and trying to spoof the analyzer into thinking it is an enhanced response cal).  If I were going from the front panel, I would select the Cal Wizard->Ecal->Check the src/rx power cal->Select 'Defined Thru' and Enhanced Resp cal from the selection and then off I go.....

I'll try uncoupling the port powers and setting Port 2 power to something legal given the attenuation value, if that gets me over the issue, all is good.  However, it would be nice to understand what is going on since I though an EnhResp cal would not need to drive from Port2 during the cal....

The 'cal all' sounds like a great feature - I remember commenting to an Agilent rep 2 yrs ago that this would be a great feature for multiple channel configurations since it is typical that there is lots of redundant reconnections.....

Thx!

我知道我不是唯一一个一直要求Cal All Feature的人，我记得在2008年或2009年与Dara谈过这件事。我昨天在一个单位安装了新的FW并且刚开始玩它。
现在，如果我们可以摆脱那些Calsets并且能够指向.csa文件。
我要求的另一个功能是将设置参数导入/导出到.INI或.CSV文件的方法，以便快速设置尚未自动化的不同工作站（或重新创建在FW更新后未加载的工作站）
。
这些在安全环境中也非常有用，因为我们无法提供状态信息并且只能下载文本，因此可以管理安全/不安全保存的切换（在频率和/或功率级别中使用xxxx）。
Spacecase



     以下为原文

  I knew I wasn't the only one that kept asking for the Cal All Feature, I remember talking with Dara about it back in 2008 or 2009. I installed the new FW on a unit yesterday and have just started playing around with it.

Now if we can just get rid of those Calsets and be able to point to a .csa file instead. Another feature I've asked for is a way to import/export setup parameters to .INI or .CSV files for quick setup of different stations that aren't automated yet (or to recreate one that doesn't load after a FW update). These would also be very useful in secure environments, since we cannot provide state information and can only download text, this could be managed a toggle for Secure/Unscure Save (with xxxx in the Frequencies and/or Power Levels).

Spacecase

是的，我同意。
获得一个包含所有信息的大“全局校准”以打开所有频道的校准将会很不错。
这可能会为项目增加一年或更长时间。
基于文本（XML？）的保存/回忆是一个有趣的想法。
我们中的一些人已经考虑过“模板”文件的想法，该文件包含许多状态信息，这些信息可以用作设置频道的起点......类似于用户预设的XML形式。
唯一的问题是它可能会很慢。
Daras有一些关于创建设备特定设置的非常好的想法，如果他可以通过组织推动它们。
您是否尝试过使用.sta文件（无校准，只是设置），然后您在之后更改频率？
然后，您可以将calsets标记为这些新频道。
顺便说一句，通过Cal All所需的更改，您现在可以在标准通道中执行引导功率校准，并在GCA通道，SMC通道和IMSpectrum通道上使用它。
只要开关和衰减器设置相同，它就会有效;
如果不是，则误差仅为0.25-0.5 dB。



     以下为原文

  Yup, I agree. It would be nice to get a big "global cal" that contains all the information to turn on calibration for all the channels. That would probably have added a year or more to the project.

Text based (XML?) save/recall is an interesting idea. Some of us have thought about the idea of a "template" file that contains lots of state information that could be used as a starting point for setting up a channel...kind-of an XML form of user preset.  The only issue with that is it would likely be pretty slow.

Daras has some very good ideas about creating device specifc setups, if he can just push them through the organization.  

Have you tried starting with a .sta file (no cal, just setup) where you then change just frequencies after?  Then you can tag calsets to these new channels.

By the way, with changes needed for Cal All, you can now do a guided power cal in a standard channel, and use it on a GCA channel, and an SMC channel, and on a IMSpectrum channel.  It will be valid as long as the switch and attenuator settings are the same; and if not, the error would only be on the order 0.25-0.5 dB.

试着结束原始问题：@Joel博士 - 谢谢。
未耦合的端口电源和设置P2合法功率范围就可以了。
我现在可以在没有任何不平等问题的情况下执行引导式校准。
但是，我仍然想知道为什么被引导的cal坚持驾驶P2。
我认为它必须是功率校准的一部分 - 如果我取消选中功率校准盒，即执行'标准'增强响应Ecal，我只看到P1驱动。
请注意，自动定位功能为OFF。
如果有必要，我可以发送一个代码片段来说明问题....谢谢你



     以下为原文

  To try and close out on original issue:

@ Dr Joel - thanks.  Uncoupled port powers and setting a P2 legal power range did the trick.  I can now execute a guided cal without any unlevel issues.

However, I would still like to know why the guided cal insists on driving P2.  I think it must be part of the power cal - if I uncheck the power cal box i.e. perform a 'standard' enhanced response Ecal, I only see P1 driven.  Note that the auto-orient feature is OFF.

If necessary I can send a code snippet to illustrate the issue....
Thanks

andy

它只是在Cal期间驱动P2还是在Cal之后也会发生？
我有一个类似的问题，我们设置了一个由端口3驱动的LO的NFx测量，以改善NF测量我通过端口3的后面路由外部源并重新校准，但原来的Port 3 Source Cal仍然存在
保持内部源，导致错误。
Joel博士，我没有考虑标准存储的文本（XML？）格式，但作为导入和导出参数的可能方法。
然后可以将它们放入电子表格中，然后将它们复制到新行并更改一些参数，然后将该行保存到新文件中。
使用Cal All，您可以简单地复制X行数并根据需要更改通道编号，频率和其他参数。
注意：帮助文件未说明所有通道（测量期间）的设置必须相同（如果有），如果已知这些设置可以轻松验证，或者甚至在单独的（常见？）部分中。
Spacecase编辑：Spacecase于2012年9月28日晚上11:03



     以下为原文

  Is it only driving P2 during the Cal or does it also happen after the Cal?

I had a similar problem where we had set up an NFx measurement with the LO driven by Port 3, to improve the NF measurement I routed an external source through the back of Port 3 and recalibrated, but the original Port 3 Source Cal was still there keeping the internal source on, causing an error.

Dr Joel,

I wasn't thinking of the text (XML?) format for standard storage, but as a possible method to import and export parameters. These could then be put into a Spreadsheet, then copy them to a new row and change a few parameters then save that row to a new file. With Cal All you could simply copy X number of rows and change the channel numbers, frequencies, and the other parameters as necessary. Note: The Help file does not state what settings must be the same for all channels (during measurement), if any, if these are known they could be easily verified, or even in a seperate (common?) section.



Spacecase

Edited by: Spacecase on Sep 28, 2012 11:03 PM

好的，所以我想出了发生了什么：当我们进行引导功率校准时，我们校准所有源和与所选端口相关的所有接收器。
Port1 ref使用端口1功率计cal校准，然后S参数将其传输到所有端口上的所有其他接收器。
但我们需要最后一次扫描来测量Ref 2上的源2功率，以了解端口2的源功率cal。这可能会导致额外的反向扫描。
我将不得不考虑在ERC期间不执行此操作是否有意义...在SMC的端口3源功率校准：如果您不选择LO校准，则不应该校准。
如果确实需要LO端口校准，可以进入电源和衰减器对话框并手动将端口3设置为OFF（而不是自动）。
应该解决这个问题
我们试图让正常情况下的工作顺利进行：你的两个案例都不是正常的80％...你很特别，所以你有机会使用特殊功能！编辑：Dr_joel于2012年9月28日
晚上11:47



     以下为原文

  Okay, so I figured out what's going on:

When we do a guided power cal, we calibrate all the sources and all the receivers associated with the ports selected. Port1 ref gets calibrated with the the port 1 power meter cal, then the S-parameter transfers that to all the other receivers on all the ports.  But we need a final sweep to measure the source 2 power on Ref 2 to know the source power cal for port 2. That is likely causing the additional reverse sweep.  I'll have to think about whether it makes sense to not do it during ERC...

On the port 3 source power cal for SMC: If you don't select LO cal, it shouldn't cal. If you do need the LO port cal,you can go to the power and attenuators dialog and manually set port 3 to OFF (instead of Auto).  should fix that.

We try to make things work smoothly for the normal case: both of your cases are not in the normal 80%...you're special, so you get the opportunity to use special features!

Edited by: Dr_joel on Sep 28, 2012 11:47 PM

Joel博士，对于端口3 cal，我是否更正它附加了新的LO源cal而不是覆盖旧的来源（端口3作为LO源），原始端口3源cal导致此问题并不明显。
我的解决方案是使用Manage Cals删除cal然后重新校准。
这是我建议参数导出功能的原因之一，它很容易（或至少更容易）发现这样的问题。
将IMD cal应用于IM频谱进行单音测量时会出现类似的问题，当我们重新应用calset时（这是我们使用calsets的少数几次之一，只是因为我们被迫）它会打开第二音并且
我们将Prt1 Src2设置为OFF。
此外，在更改其他IM Spectrum设置后，有时我们会发现Src2已再次打开。
我们使用IMD和IM Spectrum的唯一原因是IMx Spectrum没有输入音调和LO的独立控制（如果我们改变输入音调，它会改变LO）。
所以我们必须拉动Power Cal From和IMDx Cal（我仍然不明白为什么LO Source Cal需要为3阶cal的每个频率返回8个点，这只会增加传输的redundent数据的数量，更多
对于更高阶的cals ...），加载校准的IM Spectrum并使用GPIB passthrough控制外部源。
我们有几个测试（双音，单音和无输入）我们将DUT调谐到特定频率，然后我们测试其他可能的DUT输入频率（单对和双对，无输入）用于杂散输出
IF频段。
我认为使用Swept IMDx方法或IMx Spectrum（具有独立LO控制）Spacecase应该很容易



     以下为原文

  Dr Joel,

For the port 3 cal, am I correct that it appended the new LO source cal instead of overwriting the old one (Port 3 as LO source), it was not obvious that the original port 3 source cal was causing this problem. My solution was to use Manage Cals to delete the cal and then recalibrate. This is one of the reasons I was suggesting the parameter export feature, it would be easy (or at least easier) to spot issues like this.

There is a similar problem when applying an IMD cal to IM spectrum channel for Single Tone measurements, when we reapply the calset (this is one of the few times we use calsets, only because we are forced to) it turns on the second tone and we will have set Prt1 Src2 OFF. Also, there are times after changing other IM Spectrum settings that we will find that Src2 has been turned on again.

The only reason we are using IMD and IM Spectrum is IMx Spectrum does not have independent control of Input Tones and LO (If we change the Input Tones, it changes the LO). So we have to pull the Power Cal From and IMDx Cal (I still don't understand why the LO Source Cal needs to return 8 points for each frequency for a 3rd order cal, that just increases the amount of redundent data being transferred, more for higher order cals...), load the calibrated IM Spectrum and control the external source using GPIB passthrough. We have several tests (Two Tone, Single Tone, and No Input) where we tune the DUT to a specific frequency, then we test with the other possible DUT Input frequencies (Single and Pairs, and with No Input) for Spurious output in the IF Band. I would think this should be easy with either a Swept IMDx method or IMx Spectrum (with Independent LO control)

Spacecase

你有一个不寻常的案例（你通过端口3路由另一个来源），但实际上，可能有50％的客户有“不寻常的情况”（至少我参与其中）。
在重新阅读您的帖子并考虑它时，您可能使用独立的源功率校准来获取外部源。
因此它有自己的校准与其源名称相关联（称为ExtSrc1）。
因此，Port3源（内部）有源cal，ExtSrc1有源cal。
明确关闭未使用的源应该有效。
我假设您通过端口3路由内部源，以便能够在R3接收器上读取其值？
否则为什么不直接去？
IM频谱是另一个问题。
这个特殊的功能可以说是“我的宝贝”，它几乎没有进入。流行的智慧是“我们已经扫过IMD，哪个更快，为什么我们需要频谱模式？”。
（因为它很酷，这就是原因）。
因此，当它被创建时，它被狭隘地定位于IMD测量并且与源相关联。
这就是为什么当你改变东西时，固件数字“很好，一个信号源被关闭，这是一个IMD测量，我最好把它重新开启”。
您正在使用它进行正常的SA支线测试，这很好（恕我直言）。
所以这里有一些建议：不要使用转换器IM频谱，只需使用常规IM频谱。
它允许您查看输入和输出，并且由于您使用的是“线性”扫描类型，因此您可以根据需要设置频率。
应用程序将尝试在任何更改后重置IF BW，因此设置为最后一次。
使用电源/衰减器关闭第二个源通常可以工作，但您可能必须最后这样做。
如果扫描类型是线性的，我们正在考虑进行更改以关闭所有自动耦合，我认为这会帮助你。
不知道是否/何时会来，但你不是唯一一个遇到这个问题的人。
敬请关注。
此外，对于新代码中的IMS，它可以使用任何B响应cal。
因此，只需对who范围进行宽带引导功率校准，并将其保存为用户校准，然后将其应用于IM频谱。



     以下为原文

  You have an unusual case (where you route another source through port 3) although in truth, maybe 50% of our customers have "unusual cases"  (at least the ones I get involved in).

In re-reading your post, and thinking about it, you probably used the stand-alone source power cal for the external source. So it has its own calibration tied to its source name (call it ExtSrc1).  So there is a source cal for Port3 source (internal) and a source cal for ExtSrc1.

Explicitly turning off sources that are not used should work.

I presume you are routing the internal source through port 3 to have the ability to read its value on the R3 receiver? Otherwise why not go direct?


IM spectrum is yet another issue.  That particular feature is a bit of "my baby" so to speak, and it barely made the cut going in.  The prevailing wisdom was "we have swept IMD, and which is faster, why do we need the spectrum mode?".  ('Cause it's so cool, that's why).  So when it was created, it was narrowly targetted at IMD measurements and coupled closesly with the sources.  That's why when you change things, the firmware figures "well, one source is turned off, and it's an IMD measurement, I better turn it back on". You are using it for normal SA spur testing, which is fine (IMHO).

So here are some recommendations:  Don't use Converter IM spectrum, just use regular IM spectrum.  It allows you to look at both inputs and outputs, and since you are using the "linear" sweep type, you can set the frequencies whereever you like.  The app will try to reset the IF BW after any change, so set that last. Using the power/attenuators to turn off the second source usually works, but you might have to do that last too.

We're thinking of making a change to turn off all the automatic coupling if the sweep type is linear, I think that would help you out.  Don't know if/when it will come but you're not the only one who's run into this issue.  Stay tuned.

Also, for IMS in the new code, it can use any B response cal. So just do a wideband guided power calibration over the who range and save that as a user calset, then just apply it to IM spectrum.

Joel博士，我明白你的意思了，没有办法告诉PNA-x“外部源是通过端口3路由的”。
通过端口3路由它允许我们执行LO VSWR测量，只需加载不同的校准（S33），而无需在每个温度下手动重新配置设置。
我对你的IFBW参考感到困惑，早在我发现这个设置隐藏在IM频谱菜单（AVG Hard Key，More，IF BW ......）的深处，但它没有效果，每次打开时重置为1Hz
。
RBW根据所选的跨度确定点数（并且最大/最小RBW由跨度确定），这决定了扫描速度。
与频谱分析仪相比，测试时间增加了2-3倍，我正在考虑使用在端口2后面路由的MXA，以及关于是否/如何工作的任何建议。
是否有另一种方法可以提高测量的本底噪声和扫描速度，而不会出现接收器产生的杂散或压缩问题？
在较低的频率下，由于耦合器的滚降，这是否也是一个问题？
哦，我忘了提到，我第一次在安装新固件后选择了IM Spectrum时，会弹出一条错误信息，例如“Missing IF ...”，但它消失的速度太快，无法阅读。
我重置了仪器并再次选择了IM Spectrum，但它只发生过一次。
此外，我们设置了一个用户预设来保护DUT（-85 dbm和Hold），选择任何Measurement类会导致分析仪开始使用该模式/类的默认功率电平进行扫描，有没有办法防止这种情况发生。
Spacecase



     以下为原文

  Dr Joel,

I see what you mean,  there is no way to tell the PNA-x "The External Source is routed through Port 3." Routing it through port 3 allows us to perform the LO VSWR measurement, just by loading a different cal (S33), without having manually reconfigure the setup at each temperature.

I'm confused by your IFBW reference, early on I found this setting hidden deep in the IM spectrum menus (AVG Hard Key, More, IF BW...), but it has no effect and resets to 1Hz each time it is opened. RBW determines the number of points based on the Span selected (and Max/Min RBW is determined by Span) which determines sweep speed. The test times have increased 2-3x compared with the spectrum analyzer, I am considering using an MXA routed out the back of port 2, any suggestions on if/how this could work. Is there another way to improve the noise floor and sweep speed for this measurement without having issues with receiver generated spurs or compression? At lower frequencies, is this even an issue due to coupler rolloff? 

Oh, I forgot to mention, the first time I selected IM Spectrum after installing the new firmware, an error message popped up with something like "Missing IF...," but it disappeared too quickly to to read. I reset the instrument and selected IM Spectrum again, but it only happened once. 

Also, we have set up a User Preset to protect the DUT (-85 dbm and Hold), selecting any Measurement class causes the analyzer to start sweeping with the default power levels for that mode/class, is there a way to prevent this.

Spacecase

它可能是“丢失IF文件”。
当我们进行IM频谱时，我们会进行特殊的多重IF测量以去除图像，这要求IF针对不同的频率进行校准。
较旧的PNA没有IF校准，因此第一次打开IM光谱时，它会进行校准。
当我说IF BW时，我应该说“分辨率带宽”;
抱歉。
在设置其他所有内容后，您必须最后设置分辨率带宽。
试试看，看看它是否达到你需要的速度。
当然，存在速度和分辨率BW的折衷，并且由于输入耦合器，PNA与SA相比具有12-15的高噪声基底。
如果您需要较低的本底噪声（当然SA没有输入耦合器），您可以反转耦合器以将其移除。



     以下为原文

  It was probably "missing IF cal file".  When we do IM spectrum, we do special multi-IF measurements to remove images, and that requires that the IF be calibrated for different frequencies. Older PNAs didn't have the IF cal, so the first time you turn on IM spectrum, it does the calibration.

When I said IF BW, I should have said "resolution bandwidth"; sorry.

You have to set the resolution bandwidth last after setting everything else.

Try that and see if it gets to the speed you need.  Of course there is a tradeoff of speed and resolution BW, and also, the PNA has 12-15 high noise floor compared ot an SA, due to the input coupler.  You can reverse the coupler to remove that if you need a lower noise floor (of course the SA doesn't have an input coupler).

意识到我劫持了这个线程，将我的响应与我的NFx崩溃线程结合起来。
Spacecase



     以下为原文

  Realized I highjacked this thread, combined my responses with my NFx crash thread.

Spacecase