yin, gin & pez@arg
Thank you for your gracious response and compliance to my request.
In my efforts to understand your project, I think I may have had a bit of an epiphany that may start me toward understanding your needs.
I hope my response here will be of some help toward getting what I believe you are looking for from the NanaVNA users on this forum and possibly elsewhere as well.
I will begin by first referring you to page 12/15 of your presentation document:
I believe the box on the right of page 12/15 describes a four step process that is the functional flow through the software tools you have developed, and have been describing, To summarize the steps, I interpret them as follows:
1)The configuration data for the VNA measurements is entered via the (Physical State) GUI illustrated on page 4/15.
This GUI tool looks like it performs the identical function as the NanoVNA configuration setup when connected to a PC. There is no provision in this GUI tool for the input of Calibration Standards correction coefficients. If my understanding is correct, the omission of calibration coefficients is intentional. Please comment on this and correct me on this assertion. I sense this is a point you wish to make visible without ambiguity or confusion.
2) The results of four VNA measurement sweeps of a short circuit (SH.SC), an open circuit (OP.SC), a load standard (LD.LD), and a device measurement (ME.ME) are collected as 4 independent files and formatted to the input specifications of the REGION tool. These are the tasks of the ANALYSE tool.
A fifth file (INPUT.TXT) is externally and independently created which describes the magnitude only uncertainties to be used by the REGION tool. It is not clear how the uncertainty parameters are determined or specified, but they do not appear to be standards coefficients and are described as uncertainties in magnitude only. I do however suspect that they are intended to be used to statistically bias the measurements of arbitrary calibration devices (equally to or as an alternative to the use of characterized standards); and these are used for the computation of error regions surrounding each of the measured (raw) complex data points. These assertions also need corrected or made clear to remove ambiguity and enhance understanding.
3) Utilizing the output created by the ANALYSE tool; the REGION tool computes DER's and DEI's and outputs the results in both rectangular and polar form for each of the measurement frequencies.
4) Finally; the DERDEI tool formats and plots the computed output results of the REGION tool which then yields the graphical presentation that reveals the bounded region of uncertainty of the measurements.
My comments:
What I think I have been able to comprehend to this point may be close enough for me to offer a few comments that may prove useful toward reaching your goal. While it still remains unclear; I will state your goal here as a desire to obtain multiple random test cases using your software to estimate VNA measurement uncertainty. My guess is that the randomness of measurements made with uncharacterized DIY calibration standards is of special interest/benefit to you in the evaluation and defense of your software and associated algorithms.
If the software flow is as I have described, there may be some minor revisions you could make which would greatly simplify and expedite getting the data you are pursuing.
Referring back to step #1 above; the (Physical State) GUI.Input tool looks redundant to the tasks performed by VNA users traditionally and generally. It appears however that it outputs data in a format that satisfies the input requirements of your ANALYSE tool. There are many sources of S-parameter data, and the RF Test equipment industry has adopted, or at the very least supports Touchstone; a simple, text readable, defacto standardized approach to formatting, porting and saving S-parameter data. I have attached a copy of that standard here for your reference. A simple one line header (e.g. # MHz S MA R 50 ) provides all the information required to interpret the data that follows.
If your REGION tool can be easily modified to accept Touchstone formatted data as its input, you could do away with the GUI.Input and ANALYZE tools altogether, since their task is simply to gather and format the data used in the REGION and DERDEI tools. More importantly; you would gain easy access to more data than you will want or need, as libraries of Touchstone formatted S-parameter data exist in abundance. Furthermore; the ANAlYSE tool is also used to gather data via a GPIB iinterface. This too is not required, as it is only the NanoVNA's (or any VNA) data that is of interest to you, and GPIB is utilized in complex test or volume manufacturing environments, and rarely (if at all) in individual user applications.
Additionally; If modified as described, the REGION tool could be easily combined with a simple input GUI that enables a user to input the four input data files and the parameters necessary to compose the uncertainty limits file.
Appending the DERDEI tool to the REGION tool and compiling them as a single executable would result in a compact and easy to distribute and use utility that would be better positioned to gain exposure and attention across the RF test and measurement community.
Alternatively; your REGION tool if made Touchstone compatible, is sufficient (without need for distribution) to enable you to specify the requirements of random measurements that you may desire of various component types such as Antennas, components, Filters, Amplifiers, etc, and even DIY calibration kits as needed throughout the evolution of your research; leaving your team to assess, compare, and refine the output of the DERDEI tool.
I look forward to your response and perhaps a better understanding of how or if I can advance to a level of understanding where I am able to contribute in a meaningful manner.
--
73
Gary, N3GO
