多个通道的校准怎么理解

首先要做的是 - 您的网络分析仪的型号是什么？
什么是固件版本以及您有哪些选项？
您可以通过多种不同的方式进行描述，如果我们知道您提供的硬件和软件，我们可以为您提供最佳选择。



以下为原文

first things first - what is the model number of your network analyzer? what is the firmware revision and what options do you have? there are a number of different ways you can make the measurements you are describing and we can give you the best options if we know what hardware and software you have available.

N5242A，它有Option 80和83以及Time Domain。
它没有第二个来源，所以我有一个用于驱动LO的PSG。
不工作因此我无法发布固件Rev.Edited：djpritchard于2013年11月8日上午3:08



以下为原文

N5242A and it has Option 80 and 83 along with Time Domain. It does not have a second source, so I have a PSG for driving the LO. Not at work so I can't post the firmware Rev.

Edited by: djpritchard on Nov 8, 2013 3:08 AM

要迁移，sourceId：111768



以下为原文

to be migrated, sourceId: 111768

daras - 谢谢你的回复。
就这样我很清楚，如果没有升级固件，就需要为每个通道运行一个Source Power Cal。
我不完全确定你提到的细分。
您可以将一个频率的功率扫描与下一个频率的功率扫描等结合 - 从而将其保持在一个通道中吗？
而且，我只是第一次花在PNA上的时间，但是我只是在进行转换增益压缩测试，我是否必须执行SOLT校准，看到它是SMC？
我肯定会接受你购买这本书的建议。
书中是否还有编程示例。
再次感谢编辑：djpritchard于2013年11月8日下午2:49



以下为原文

daras - thanks for your reply. Just so I'm clear, without the upgrade to the firmware, it would be a matter of running a Source Power Cal for each channel. I'm not exactly sure about your mention of segments. Can you tie a power sweep at one frequency to a power sweep at the next frequency, etc., etc - thus keeping it to one channel? And again, I'm just getting to spend some time on the PNA for the first time, but is I'm just doing a Conversion Gain Compression test, do I have to perform a SOLT cal, seeing it's a SMC?

I'll definitely take your suggestion on purchasing the book. Are there also programming examples in the book.

Thanks again

Edited by: djpritchard on Nov 8, 2013 2:49 PM

再次 - 不知道你正在运行什么固件版本，我不知道你的盒子有什么功能。
但是，SMC通道始终有自己的校准向导。
因此，请允许我重申_ *您不需要使用SMC通道（或通道）* _执行单独的源功率校准或s参数校准。
当您激活SMC通道并进入Response-> Cal-> Start Cal-Cal Wizard选项时，启动的校准向导将执行功率和s参数校准。
对于功率校准部件，您需要一个功率传感器，对于S参数校准部分，您需要一个ECal或机械校准套件，与您的DUT /电缆的连接器相匹配。
所以我们假设您需要对上变频器进行压缩测量，其中RF从1-2 GHz扫描，LO固定在10 GHz，因此IF从11-12 GHz扫描（并且您没有选项
086）。
还可以假设，对于压缩测量，跨越跨度需要21个频率点，您需要将-50 dBm的RF功率扫描到0 dBm，步长为0.5 dB。
这是51个功率点，21个频率点和总共1071个测量点。
端口1上的PNA-X源可以从-30 dBm扫描到大约+10 dBm，源衰减器设置为0 dB。
因此，您应该创建一个SMC通道设置，以21个频率点扫描RF /固定LO，并将RF功率设置为标称-5 dBm。
然后，您可以使用SMC校准向导校准此通道，并将校准值保存为“用户校准”。
这将是您的主校准。
现在您有两个选择：您可以将此通道复制50次（总共51个通道），并在每个通道中将RF功率设置为-25到0 dBm的值，步长为.5 dB，或者您可以复制
通道20次（总共21个通道），在每个通道中，您可以将扫描类型更改为扫描功率，RF / LO / IF设置为固定，以及您需要的21个RF频率之一并设置开始
/将RF功率停止在-25和0 dBm，通道的点数设置为51.在任何一种情况下，所有通道都指向您之前创建的相同主校准集。
然后，您可以扫描所有通道，并且您将拥有在每个频率上计算压缩所需的1071个数据点。
如果我在你的情况下我会选择第二个选项（21个通道），因为当通道设置为电源扫描并且我显示SC21作为输入功率的函数时，我可以简单地在每个SC21跟踪上放置压缩标记
读出每个频率的输入和输出压缩功率。
如果您选择第一个选项（51个通道），则必须将所有数据拉出并对其进行后处理以计算压缩点。
现在，使用分段扫描模式，您可以复制具有51个相同段的51通道场景，其中每个段具有不同的RF功率级别。
通过编程和手动设置分段情况稍微复杂一点，但是在整个测量时间内你会得到一点点优势，因为你只需要扫描1个通道而不是51个。数据采集的总数是相同的
，但是当您从一个频道切换到另一个频道时，固件会有一些开销，而在分段情况下，您不会产生这种开销。
* + _ Comercial Break _ + *选项086 - 增益压缩应用程序消除了我刚才解释的所有复杂性。
您将使用一个通道来设置频率范围和功率扫描范围，您将进行一次校准，所有数据最后都在一个通道中，使用SMART Sweep选项，您甚至无需全部收集
每个频率有51个功率点（每个频率只需3-4次功率测量就可以非常准确地计算出压缩点），因此最终测量速度比完整的1071数据采集快得多。
并且在1个通道中编程要比51,21或51个段更容易。
* + _ Comercial Break结束_ + *遗憾的是，本书没有编程示例，但是您可以在此论坛中找到很多编程示例，也可以在_PNA在线帮助中找到。



以下为原文

again - not knowing what firmware revision you are running, I don't know what capabilities your box has. However, the SMC channel has always had its own calibration wizard. So let me reiterate _*you DO NOT need to perform separate source power calibrations or s-parameter calibrations with an SMC channel (or channels)*_. when you have an SMC channel active and you go under the Response->Cal->Start Cal-Cal Wizard option, the calibration wizard that is launched does a combination of power and s-parameter calibrations. for the power calibration part you will need a power sensor and for the s-parameter calibration portion you will need either an ECal or a mechanical cal kit that matches the connectors for your DUT/cables.

so lets assume that you need to do compression measurements for an upconverter where the RF is swept from 1-2 GHz, the LO is fixed at 10 GHz and the IF is therefore swept from 11-12 GHz (and you don't have option 086). Also lets assume that for your compression measurements you will need 21 frequency points across the span and you need to sweep the RF power from -25 dBm to 0 dBm with 0.5 dB steps. so that is 51 power points, 21 frequency points and a total of 1071 measurement points. the PNA-X source on port 1 can sweep from -30 dBm to about +10 dBm with the source attenuator set to 0 dB. so you should create an SMC channel set to swept RF/Fixed LO with 21 frequency points and the RF power set to a nominal -5 dBm. you can then calibrate this channel, using the SMC calibration wizard and save the cal to a "User Calset". this will be your master calibration.

now you have two options: you can either copy this channel 50 more times (for a total of 51 channels) and in each channel set the RF power to a value from -25 to 0 dBm with .5 dB steps or you can copy the channel 20 more times (for a total of 21 channels), where in each channel you change the sweep type to swept power with RF/LO/IF set to fixed and to one of the 21 RF frequencies that you need and the set the start/stop RF power to -25 and 0 dBm with the number of points for the channel set to 51. In either case all the channels are pointing to the same master cal set you created earlier. you can then sweep all the channels and you'll have the 1071 data points you need to compute compression at each frequency. If I was in your situation I would choose the 2nd option (21 channels) because when the channels are set to power sweep and I am displaying SC21 as a function of input power, I can then simply put a compression marker on each SC21 trace and read out the input and output compression power for each frequency.

if you choose the first option (51 channels), you'll have to pull all the data out and post process it to compute the compression points. now, with segment sweep mode, you can replicate the 51 channel scenario with 51 identical segments, where each segment has a different RF power level. setting up the segment case is a bit more complicated programmatically and manually, but you do get a slight advantage in over-all measurement time, because you will only need to sweep 1 channel instead of 51. the total number of data acquisitions is the same, but there is some overhead in the firmware when you switch from one channel to another and in the segment case, you don't incur that overhead.


*+_Comercial Break_+*

option 086 - Gain Compression Application removes all the complexities that I just explained. you'll use one channel to setup the frequency range and the power sweep range and you'll do one calibration and all the data is in one channel at the end and with the SMART Sweep option, you don't even have to collect all 51 power points at each frequency (it figures out the compression point very accurately with just 3-4 power measurements at each frequency) so the final measurement is a lot faster than doing the full 1071 data acquisitions. And it is a whole lot easier to program in 1 channel than 51, 21, or 51 segments.

*+_End of Comercial Break_+*

unfortunately, the book does not have programming examples, but there are a lot of programming examples that you can find in this forum an also in the _PNA online help_.

daras - 感谢最后的回复 - 现在我已经有了很好的理解。
我将努力在将来添加Option 86，但就目前来说，这是关键时刻，我没有时间去反对bean计数器。



以下为原文

daras - thanks for that last reply - now I've got a very good understanding. I'll work on getting the Option 86 added in the future, but for now, it's crunch time and I don't have time to go up against the bean counters.

很高兴我能得到一些帮助。
仅供参考，您可以随时联系您的安捷伦代表，并为我们的任何纯软件选项（如086）申请免费试用许可。
有时显示bean计数器速度和设置简单的实际差异是放松钱包的技巧:-)让我们知道我们是否可以提供更多帮助。



以下为原文

glad I could be of some help. FYI, you can always contact your Agilent Rep and ask for a free trial license for any of our software-only options (like 086). sometimes showing the bean counters the actual difference in speed and setup simplicity does the trick in loosening the purse strings :-)

let us know if we can be of any more assistance.