NanoVNA Custom Software

[joeqsmith]

Re:  Regrettable you don't want to organize the help you've already provided.
People will not read it which is indeed regrettable.

Notice how every picture you show the sweep is turned off.   Surely you don't expect it to sweep when you are telling it not to.  We can see in the last screen shot that the VNA is not sending any data which I would expect when the sweeps are turned off.   

Lets assume this is not the problem.   For me to look into it, you need write a detailed procedure that I can use to replicate the problem.  Document every step from the time you turn on the VNA.   If you can do this, there is a good chance that if there is a problem, I could fix it or at least tell you where the problem is.   Consider that I used the V2Plus4 a fair amount with this software and had no problems like you describe.
***
Make sure you include the firmware information as well.  Every detail is important. 

[joeqsmith]

Looking at the code, the only thing that make sense is not enabling the sweep.   You could run through the calibration process with sweep turned off.  The software will not care and happy do it.   Of course, it will be very fast as its not collecting any data.  The coefficients will not be created like you show as there was no data collected to create them.  The cal LED will be active as it made it through the calibration.   

However, you wrote:

I'm getting nothing when I sweep. Seems locked up.

I suspect you actually do press the sweep after you have ran through a cal without the sweep enabled.  The software being cal'ed is of course calculating the display values based on these nulled coefficients.   It's going to display a lot of nothing just like you asked.  You then turn off the calibration and like magic, it starts displaying data because it no longer uses the nulled coefficients. 
 

12.5   Displaying Data

The software will default to a center frequency of 10MHz with a 2MHz span and 201 data points (Frequency spectrum shown in Figure 11).  The V2+ was programmed with these defaults when the connection was made.   The V2+ is actually sweeping at this time. Select the Sweep button and the software will begin collecting data from the V2+. 

Maybe there is actually a problem but you have not provided any details on how to replicate it.   I suspect like many, it's easier for you to post on the forums than to use that time to try and understand what is going on.  Most of the basics were covered during the videos and the manual.   

[jspencerg]

I'll retry and document procedure today.

[jspencerg]

Re: "freezing up"
Realized as soon as I read your comments: I was not enabling sweep BEFORE calibrating. Works fine. Duh. Newbie mistake.
Calibrated 1001 points from 1M to 4.5G. Connected said same 20db attenuator.  Lots of jitter at higher frequencies,  but looks as though average would be fine till around 3.25G. Shown is results using normalization. Not appreciable difference without normalization.
Thanks for your help!
I've been salvaging components. Interesting in characterizing Inductors and common mode transformers. Lots of approaches from you and others.  Going to build holder(s) to test.  Fun stuff!
Parts for transfer relay arriving within week.  Need to study up on circuit build to power and connection details for usb. I'm so ignorant.

[joeqsmith]

It looks like you may have the leakage term disabled.   OWO mentioned the problem with the V2Plus4 when I first received mine.  I would just save that waveform to memory then enable the leakage and see how much it changes....

You seem to failed to read my post so I will repeat myself, 

I suggest you look at the specs for your V2Plus4:  https://nanorfe.com/nanovna-v2.html

And again, READ the attached. 

Surely you understand that 4.5GHz is > 4.4GHz??

***
It also appears that you calibrate the system and then use the normalize after the cal. 

[jspencerg]

Re: Your feedback
1. Thanks for reminder about leakage. I'll study that and incorporate.
2. I know I'm exceeding freq. limits.  I'm pushing beyond to see if the published limit is accurate for my unit.  No harm there.  Will not try that with dbm limit!
3. Yes.  I calibrated before trying normalizing.  I did not realize that would be a mistake.
Thanks.

[joeqsmith]

I assume you ran your V2Plus4 standalone, setting the range to 4.5GHz and it allowed it.   From that you feel there may be something to gain.  My software wouldn't care and assumes you know what you are doing.   The reason I mention it is I think the firmware I have for the V2+4 will coerce the data to stay within the limits.  Obviously, if that were the case you are just fooling yourself.  One way to get more BW is to go with the LiteVNA.  It's about half the cost and offers a lot more bang for the buck anyway.   I'm surprised that you wouldn't have gone with it.     
 
I don't consider attempting to run your equipment outside of the specified range,   normalizing after cal  or even running a calibration with the data collection turned off a mistake.  Again,  I try to keep the software as open and flexible as possible as it is an engineering tool after all and not recommended for the beginner.   Your mistake comes when you suggested the software locks up when it's just doing what you told it.   One way to avoid that would be to read the manual.     

For the leakage, just turn it on.  There is no need to recal or change your setup.

***
I just checked my V2+4 with the original firmware (20201010-86c7055) installed and indeed mine will coerce the setting to keep it in range.  If I enter 4.5GHz, mine will coerce to 4.4GHz.  So you must have newer firmware than mine.   Or you believe my software somehow overcomes this limitation, which of course, it does not.     

[joeqsmith]

Solver64 is now available.   

[realfran]

SOLVED ( ini.file) 

[cyp_eev]

Solver64 is now available.   

Works perfect, no problems so far.
Many thanks!

[jspencerg]

Hello Joe,
"the LiteVNA.  It's about half the cost and offers a lot more bang for the buck anyway.   I'm surprised that you wouldn't have gone with it."
I asked you which version of the vna's to get.  You said either the original or the v2plus4 would be good for learning.

"read the manual"  I did.  I still make mistakes.  This is all new I did not have the common sense that the calibration session was going too fast (without first activating the sweep).  Details, details.  Still working on the details. 

I found a good reference for calculating the Z (and other associated) parameters.  Studying the math of it.  I don't suppose I could look at the code you settled on for your calculations?

[2x2l]

JoeQSmith:

https://www.ni.com/en-us/shop/labview/select-edition/labview-community-edition.html


https://www.ni.com/en-us/support/documentation/supplemental/20/labview-community-edition-usage-details.html

In general, non-commercial and non-industrial references any use case that is not intended to make a profit. The LabVIEW Community Edition is intended for personal projects. Here are some examples:


Home hobbyist projects
Making free, open-source projects or add-ons for the community to reuse
Using LabVIEW to study for an upcoming certification
Using LabVIEW at home to keep skills sharp or try out new ideas outside of work
[/b]

I'm not particularly sure how restricted (if at all) the community edition is, but you certainly do fit at least one of those. 'add-ons for community use'. I'm one of those weird people who reads through the whole EULA, and it doesn't have anything about redistribution of software, but based on what's specifically stricted by the EULA, you're definitely fine on that end. I'd feel safe using it if I were in your position.

But other than that, you should start doing things along the lines of "Marco Reps" but someone with +25 years. The content is technical yet diverse, funny delivery and good content. He manages to make the concept of guarding against current loops on test equipment funny. A feat most people would think impossible. He made his own milling machine and got like 10 videos out of that-- testing different controller setups and things like their acceleration profiles, different ballscrews to see how linear they are, spindles and their run-out, and then finally showing the project & it's performance.

I think he got reliability works down to the mid-nanometers. Sure, Edmunds would build you something with the same numbers but enjoy paying a few million for it. He got way past metal machining numbers, competing with optical numbers (where you break out optical test flats under HeNe(600ish nm?) with eh, 8 interference fringe patterns gets you into the double digit nanometers. (He obviously can't *do* optical because it's not equipped with the necessary spindles, vibration damping, tooling to avoid subsurface fracturing, etc, and he was definitely hitting those numbers against standard cold rolled steel.) 

Wow that went off topic.



Again, for 22000 subs (crazy), you've tested like...them all. Except the Euro version of the UT61E. I forgot where I heard this mentioned, but they're supposed to be hardened up. I think using an actual HRC fuse and some such? If you do test it, and it turns out to be a good buy, PM me 24 hours before you post the video, so I can buy one before the rush hits.

Testing components too would be really interesting. Get a bag of GDTs from different manufacturers, and compare them. Maybe from different lots, and see how much conservative they play with their spec sheet numbers. There are some vendors (TDK? Epcos? Vishay almost certainly..) that have the MOV/PTC in one package and see what they can handle, their failure modes etc. Educational for the community and blowing things up is always fun.


And to test them, maybe design a unit that will meet the official UL test standards.  That'd probably be one of the most interesting series you could possibly do. Have an 8 part series where 10 minutes of the content is the theory injected at arbitrary spots in the dialogue, with the remainder being "most people would do this, but they're idiots: do this instead". Or the famous issue of: One or two people put out an app note, everyone reads that app note, then 20 years down the line people are talking about 45 degrees at DC. "they're idiots, ..." Or just mentioning shit that most people don't really think about. The first chapter of Hall, Hall and McCall did this really well. with dispelling the rumors on where the energy is actually passing through (the dielectric, not ON), along with how the EM fields develop and *why*.   Though I'd be reluctant to walk through something like that on YouTube. Judging by the emails you seem to get...well...

[realfran]

Hello, I start the test of the program, the graphics line on the monitor screen hare too thick on v.3.
I put some screenshots to see the difference on v.2.08 the line got better definition and better readability on V.3 some go to mixup.

[joeqsmith]

LabView allows you to change the line colors and thickness.   I have added the ability to save the main graphs settings to the defaults file.

[joeqsmith]

Hello Joe,
"the LiteVNA.  It's about half the cost and offers a lot more bang for the buck anyway.   I'm surprised that you wouldn't have gone with it."
I asked you which version of the vna's to get.  You said either the original or the v2plus4 would be good for learning.

Perhaps a case of SISO?  You didn't provide a reference.  If you would like less vague answers, learn to ask more detailed questions. 

[realfran]

LabView allows you to change the line colors and thickness.   I have added the ability to save the main graphs settings to the defaults file.

Is only one comment for the difference from V.2.08 to V.3.0 for the better readability of V.2.08.
I'm computer illiterate 
 

[joeqsmith]

Comparing a BP filter on my PNA vs the Lite.   Different standards, cables and method were used.   

[optotester]

Hello, is that default LiteVNA or with the capacitor mod you mentioned previously ?

[joeqsmith]

Unmodified LiteVNA with custom diode dropping USB cable.

[ehcurrie]

Sure do wish I knew how to solve this problem... does lvsound2.dll have to b 8)e in a specific location?

[realfran]

[joeqsmith]

Sure do wish I knew how to solve this problem... does lvsound2.dll have to b 8)e in a specific location?

This sound DLL seems to be a bit of a bastard child.  Normally, I would expect it to be included as part of the runtime like everything else.  For what ever reason, the sound is a special snowflake.   When building the EXE, LabView places the DLL into the Data directory.  The proper directory structure is shown with Github. 

Again, you need the correct VISA and runtime for it to run and communicate with the VNA.  It's a 64-bit app and tested on Windows 10 only. 

[ehcurrie]

I have been using the correct versions of VISA and the Labview-runtime (64bit).  Actually, I have a complete install of LabView. But nontheless, no matter what I have tried in terms of placing the lvsound2.dll I can't get past the program complaining that it can't find the DLL I assume that there is no way to option out the sound support. It's so frustrating to be unable to use the app with a NanoVNA because of sound.

[joeqsmith]

I have been using the correct versions of VISA and the Labview-runtime (64bit).  Actually, I have a complete install of LabView. But nontheless, no matter what I have tried in terms of placing the lvsound2.dll I can't get past the program complaining that it can't find the DLL I assume that there is no way to option out the sound support. It's so frustrating to be unable to use the app with a NanoVNA because of sound.

I haven't seen any problems with the sound posted in the LV forums.  Sadly you decided to chop off the screen shot and have supplied very few clues to what you are doing.   I can replicate the errors you posted but I really have to go out of my way to screw things up. 

Remove the DLL. Launch the program and when it prompts for the location of the DLL, select cancel.  Tell it to continue to load anyway.   It will show an error but force it to run.  I would expect trying to run it this way would be obviously a bad idea but it will reproduce the same results.

If you want to try an solve it, you will need to supply a lot more detail about what you have done.   Otherwise, I would just use the open source program. 

[joeqsmith]

I tried to grab an older 32-bit version of the DLL and place that into the correct directory.  They have the same name.  I then attempted to load it and it will not prompt for the file location.  The error is still set and when I attempt to force it to run, the software once again shows the same errors.  Of course, no one in their right mind is going to expect my two examples to work.   

Then again, with you having the full blown LV installed, I can believe you tried to run with a DLL you had.  I have no way to know without my crystal ball.