jerseydevil Posted August 30, 2005 Report Share Posted August 30, 2005 I started looking for program examples of how best to represent multiport s-parameters and t-parameters in Labview and was surprised to find only the definitions in the NI dictionary. I expected to at least see some common 2-port s-parameter programs, but couldn't even find any of these. Is no one doing any of this, or maybe I'm just using outdated terminology? :headbang: Quote Link to comment
Michael Aivaliotis Posted August 31, 2005 Report Share Posted August 31, 2005 I started looking for program examples of how best to represent multiport s-parameters and t-parameters in Labview and was surprised to find only the definitions in the NI dictionary. I expected to at least see some common 2-port s-parameter programs, but couldn't even find any of these. Is no one doing any of this, or maybe I'm just using outdated terminology? :headbang: 5909[/snapback] Are you talking graph displays like Smith Plots? Look under the Graph palette. Quote Link to comment
jerseydevil Posted August 31, 2005 Author Report Share Posted August 31, 2005 The smith chart is a typical display function for s and t parameter arrays, but it does not imply any specific format for representing logical groups of s parameters such as the typical multiport s parameters, it just displays complex arrays. I already measure, read in, and manipulate large groups of s parameters from Network Analyzers although up until now these are mearly a collection of arrays, in some cases multidimensional arrays that are formed in an s parameter layout. The newer analyzers like the Agilent E5071B naturally operate on multiport s parameters and t parameters, so I started looking into how Labview would represent such data, and was surprized to find nothing. It is clear that direct matrix multiplication and the other matrix operations can be used on 2 dimensional complex arrays, thus implying the correct form for a multiport s parameter data structure, however Labview has no support for typical things like impedance transformation and t parameter conversion. They certainly have quite a few complex matrix math functions, but I was very surprised to see that it (at least Labview 6i) is virtually mute on the whole issue of s parameters. I don't mind cooking my own, but I think it is always better to work as a community. I was also surprised to find nothing about this in LAVA. Quote Link to comment
Michael Aivaliotis Posted August 31, 2005 Report Share Posted August 31, 2005 The smith chart is a typical display function for s and t parameter arrays, but it does not imply any specific format for representing logical groups of s parameters such as the typical multiport s parameters, it just displays complex arrays. I already measure, read in, and manipulate large groups of s parameters from Network Analyzers although up until now these are mearly a collection of arrays, in some cases multidimensional arrays that are formed in an s parameter layout.The newer analyzers like the Agilent E5071B naturally operate on multiport s parameters and t parameters, so I started looking into how Labview would represent such data, and was surprized to find nothing. It is clear that direct matrix multiplication and the other matrix operations can be used on 2 dimensional complex arrays, thus implying the correct form for a multiport s parameter data structure, however Labview has no support for typical things like impedance transformation and t parameter conversion. They certainly have quite a few complex matrix math functions, but I was very surprised to see that it (at least Labview 6i) is virtually mute on the whole issue of s parameters. I don't mind cooking my own, but I think it is always better to work as a community. I was also surprised to find nothing about this in LAVA. 5924[/snapback] I misunderstood. I thought you were interested in the display aspects of S-par values. This is why the post was moved to the user interface forum. Perhaps another move is necessary. In any case, I assume you've seen the complex palette? PJM always calls me a master of the obvious but I'm always surprised by how many people miss this palette. No, LabVIEW does not have functions to operate on multiport S-parameters. On the other hand, why should it? Are you trying to perform correction on the measurements based on calibration data taken for the multiple ports? To do this you need to apply the formulas required based on your calibration method. Most of these correction formulas are public knowledge and can be easily implemented using basic math primitives. All the primitives (+,-,/,etc) work natively on complex numbers so no special transformation is required. Some calibration methodologies are patented and can only be used after paying royalties to third parties, however those are not always necessary. BTW, are you doing a full 12-term 2 port calibration on the analyzer ports? I assume you are using some sort of switch matrix to automate the calibration and measurement process, right? Quote Link to comment
jerseydevil Posted September 2, 2005 Author Report Share Posted September 2, 2005 I'm maybe overly familiar with the complex palatte, even have used the complex linear algebra palatte. One of these days I may even work my way up to the advanced complex linear algebra palatte. I use switches and do the 2 port stuff. I thought this was pretty common stuff, and I was finally ready to join the rest of the world by looking into how Labview handles it. The idea occured to me to look after I saw how many "waveform" functions they had. Given the extensive application of s parameters and Labview's natural bias toward test and measurement, I expected a treasure trove of functions and features, not a totally mute package. Are s-parameters actually patent protected? Blame it on the 60's. And NI is too cheap to pay those royaltys? Quote Link to comment
Michael Aivaliotis Posted September 2, 2005 Report Share Posted September 2, 2005 Are s-parameters actually patent protected? Blame it on the 60's. And NI is too cheap to pay those royaltys? 5952[/snapback] No, S-parameters are NOT patented, I was talking about the calibration algorithms. As far as LabVIEW's capabilities, I have yet to find any limitations. Even in the RF\Microwave field. Are you using any automation at the moment? If so, what software are you using? Another thing to look at is NI's add-on RF Software tools for your industry. Perhaps if you explain what specific problem you are trying to solve then I can guide in the right direction. Quote Link to comment
jerseydevil Posted September 2, 2005 Author Report Share Posted September 2, 2005 I, too, don't see any limitations with Labview. I was just recently looking at USB and now I see that they support it in the newer release. NI has been excellent and Labview has provided half a career for me now. I guess it's funny that when you can find something that really works, its easy to just dig in and stay productive, and not even look up. Labview=Productive. I see from your click link that NI has become very "signal oriented". This seems like the best approach for most test engineering. Quote Link to comment
xseadog Posted September 2, 2005 Report Share Posted September 2, 2005 I, too, don't see any limitations with Labview. I was just recently looking at USB and now I see that they support it in the newer release. NI has been excellent and Labview has provided half a career for me now. I guess it's funny that when you can find something that really works, its easy to just dig in and stay productive, and not even look up. Labview=Productive. I see from your click link that NI has become very "signal oriented". This seems like the best approach for most test engineering. 5961[/snapback] Just to add my 50 Quote Link to comment
jerseydevil Posted September 2, 2005 Author Report Share Posted September 2, 2005 No driver yet, but it is only a matter of time. :beer: Enjoy the end of summer BBQ's :thumbup: Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.