NanoVNA Custom Software

[Bicurico]

@joeqsmith: Thanks for your extensive reply - much appreciated.

Kind regards,
Vitor

[hagster]

How can you tell which the good ones are?

[Bicurico]

I don't know, but I can tell you this:

I ordered the CHEAPEST one i found on eBay (https://www.ebay.de/itm/Nanovna-50KHz-900MHz-Vector-Network-Analyzer-UHF-HF-VNA-UV-VHF-Antenna-Analyzer/133112732589) and it looks like a mixture of both the black pictures: the PCB and shields look like the "best" one, while the lettering on the case is more similar (but not the same) as the "worst" one.

Note that I mean cheapest, but complete: with the "calibration kit" - which consists on three SMA caps: open, closed and 50 ohm, as whell as two pairs of SMA cables and a male-male SMA connector.

The device came in a nice plastic case, which I did not actually expect.

I think that it is just a case of someone having posted these pictures without a real explanation on what/why one devices are supposed to be better than others. The pictures are rubbish and the only visible difference is the "NanoVNA" print next to the screen, which is bolder on the "good" device and thinner on the "worst" device. I call it rubbish.

From what I have read before ordering mine is that the white ones may come without the battery and older black ones had less/no shielding inside. Also, you might be buying just the device without accessories, so make sure you get them, too.

Regards,
Vitor

[bitseeker]

The only way to know for sure is to buy it only from the originator's store, which is on Taobao. Buying from unauthorized resellers always carries some risk. You can't be sure that the photos match what they're selling on any given day or if they even understand what they're selling in order to answer your question about which one they have in stock.

The one Bicurico bought is the good one. The font matches the one on the author's github page and you can see the shielding in the pic taken of the front edge. Of course, since any reseller can "borrow" photos, it's not a guarantee that you'll get that, but those are the characteristics to look for.

[bitseeker]

I think that it is just a case of someone having posted these pictures without a real explanation on what/why one devices are supposed to be better than others. The pictures are rubbish and the only visible difference is the "NanoVNA" print next to the screen, which is bolder on the "good" device and thinner on the "worst" device. I call it rubbish.

The author could've done a better job of explaining what to look for in his comparison pics, but they're not rubbish. There are clear differences after you examine them and some eBay listings.

Glad to hear you got a good one. What does the case look like?

[joeqsmith]

@joeqsmith: Thanks for your extensive reply - much appreciated.

Kind regards,
Vitor

Glad to help. 

The current price of PSPICE is $6400USD and the student version wouldn't support i.   I'll have a look at this simulator.  If it looks like I can sort out some sort of demo, I'll go ahead and add Touchstone support.   

Thanks again.

Another option is the Microcap software which recently has become free:

https://www.eevblog.com/forum/projects/microcap-12-is-free-(as-in-beer)-now/

Thanks.  I will have a look at it as well.   

***
It does appear to support it as well.  I would think between the two versions of SPICE, I should be able to get something working.   Thanks again.   

[joeqsmith]

I tried installing both SPICE programs and started to play with them a bit.  It looks very promising so I have added the ability to import a s1p file into my Nano software and want to verify the results.   

Is anyone aware of any free tools used to process Touchstone files?   I am basically just needing something like AppCAD but am looking for something that can plot inductance (and other) directly. 

Plan is to take a component that I have a touchstone file for, import it and make some measurements.  Then test the actual part and compare the results.   Once that works, I will go ahead and add the ability to export.   Then attempt to use the S-parameters from the Nano in SPICE.   

All seems doable but not trusting my math skills. 

[malagas_on_fire]

Would any of this program suits your needs?

https://www.rfpage.com/free-tools-to-analyze-and-plot-s-parameters-file/

https://www.ag-rf-engineering.de/products/software/s-parameter-viewer/

[joeqsmith]

This part was produced by E F Johnson.    Here you can see AppCAD  showing the Touchstone file created using a slightly higher end VNA.    Also shown is the same file imported with the Nano software.  Cursors were set at 1.6MHz and 8.7MHz.   Part was installed on my BK meter as a sanity check.   The part attached to the Nano, sweeping  from 1 to 10MHz and cursors were set at roughly the same location.

Obviously, the Nano is never going to replace a higher end VNA you would use in the lab.   Still, at $50ish dollars, I am having a hard time finding a reason to bash it.  If you wanted to learn about VNAs, I still think this is one of the best investments out there. 

Odd, attachments didn't load....

[joeqsmith]

And, just for fun comparing $20,000+++  with $? (whats an 8754A worth today) and $50.   Three VNAs, three answers, just like DMMs.   

[joeqsmith]

Exporting from the Nano and reading with AppCAD.   

Time to read a few SPICE manuals.     

[joeqsmith]

I ended up installing the Micro-Cap simulator.  Like the Nano, the cost was so low (free to download), it's going to be hard to find any fault with it. 

I made up a simple circuit with the 1 port model.  I then imported a Touchstone file from the Nano fully expecting it to crash, started the simulation.    It actually converged!!   

It's way too early to say if it can spit out any useful data but the fact it's doing anything at all is impressive.  I'm keeping my expectations low.

[joeqsmith]

Just for a simple easy sanity check, I swept a 50 ohm resistor from 50K to 900MHz and stored that into a Touchstone file.  I then imported that into AppCad.  No problem.  Then I imported that same file into MicroCap and using a source with 50 ohms output impedance,  ran a sweep up to 900MHz and plotted the loss.   Looks like -6dB to me. 

I would say we have some meaningful data for S11.   The next step is to get S21 working, then add a transfer relay to the Nano.   Oh wait, the relay would cost 10X more than the Nano. 

****

I have attached the Touchstone file which would be renamed before using.  The MicroCap software is smart enough to identify touchstone files but it does not automatically calculate the number of data points. 

[tinhead]

Joe,

any plans to release the LabVIEW interface?

[joeqsmith]

Joe,

any plans to release the LabVIEW interface?

While a few people have now asked,  I have no plans to release it at this time.  The problem is lack of time to support the few people that would actually try it.   It's really what I would call an engineering tool, not what I would consider an end product.   

So while I appreciate the effort put into an optimized PC program, I wonder if this is really worth it if the measurements are not very precise anyway`

Oxdeadbeef seems very much aware of the problem.  While it's not their time, I appreciate their comment.   It's a great little unit for the beginner IMO, but just how much time do I want to invest in it.  Someone wrote me about creating something in Python and it not being much more effort.  Another wanted something for LINUX.  I wouldn't be at all surprised if we don't start seeing different software for it.

I still plan to make one last video for the Nano and will show the latest software for it. 

[joeqsmith]

Micro Cap SPICE simulation with the NanoVNA

[malagas_on_fire]

For software programmers a flowchart of you current release SW and maybe serial data protocol could get them on the road for a multi-platform GUI written on python, java, etc ... By the way the mug looks nice, but is it filled?

[OwO]

I don't understand why there are so many separate community efforts to develop VNA GUI software but none are open source, leading to fragmentation and duplication of efforts. If you are serious about this project please just post it on GitHub. There doesn't seem to be much software skill on this forum and putting the code out there gives real software developers a chance to look at it and possibly develop it further, possibly developing into a generic VNA GUI that can easily support new devices. There is nothing to be gained by keeping the source closed, it will just render your project obsolete as soon as you stop working on it. To me releasing a .exe is the same as not releasing anything at all.

[hagster]

Joe has built this for himself alone. He doesn't want or need it to be a solution for anyone else.

Besides, i doubt there are many developers that would want to jump in on a project based on LabVIEW.

I would love to see a good Open source GUI though. There is no reason it could not support multiple VNAs including the big expensive Keysight and R&S units via scpi. Maybe then they would stop charging an arm and a leg for simple features like TDR(it's just an IFFT ffs).

[_Wim_]

I must say I can understand joeqsmith choice. There is indeed always the expectation when somebody has gone through the “trouble” of compiling/installing your software and it doesn’t work, the developer chimes in to provide some support. And this can cost quite a bit of the spare free time we all have to play around. For successful apps with wide community support this gets somewhat better, but a software package like this is unlikely to get this status.  Also, like many of us we are interested in a certain topic only for a certain period, and then we move on to the next interesting subject. When making software available it is more difficult to leave the topic because there is the expectation to fix bugs and continue development.

Also, as stated correctly above, this is not a programming forum, so this means not so many people will go to the trouble to actually read the source code, most will just run the exe and hope it will work, and ask question on the forum if it doesn’t, especially for a low cost device like this.
I find some of the above comments sounding like joeqsmith is doing us a disfavour, by not posting his software online, while off course the opposite is true, by making the informative videos and comparisons with a real VNA and also showing what is possible with some additional effort. 

[radiolistener]

this is his right to not publish his source code. Some bad people may use his published source code, remove copyright and claim it like this is their own development. They can insult him and trolling just for fun. Some of them may use it in their commercial projects. And the author will loss interest for this project, will stop development and abandon it. And the project will die. I catch similar issues on several popular projects which I published as open source in the past. These things may happens when your project may affect some other commercial projects. Since you publish it for free these guys who make money on the similar projects will start to fight against you. So, now I don't publish source code of my projects which is interested for me for further development. I can publish some parts which is not interested for me anymore, but not these which I'm still working on.

Just think - he may not publish it at all. I have several projects which I never published and I know many people will be interested in it. But I will not publish even binaries of it, due to different reasons. So, if he decided to share just a binary, just say thanks to him. This is his work and he can decide what part of his work to share or not to share. This is better than nothing at all

I think joeqsmith doing good job, he sharing his experience and knowledge with other users, he even shared his tool. If you want source code to make something similar, then just make your own. No needs to blame someone for not sharing something that you're needs. There may be reason for this.

[joeqsmith]

I don't understand why there are so many separate community efforts to develop VNA GUI software but none are open source, leading to fragmentation and duplication of efforts. If you are serious about this project please just post it on GitHub. There doesn't seem to be much software skill on this forum and putting the code out there gives real software developers a chance to look at it and possibly develop it further, possibly developing into a generic VNA GUI that can easily support new devices.

If there really are separate efforts as you suggest, I would suspect it has to do with the various communities having different objectives.  Some want LINUX support, some want Python, some C#, some VB, some want it running on Apple products.    My goal was just to put some sort of demo together to help show how to use the Nano.     

I reversed the protocol and put together the main program in a couple of days.  Reading about Touchstone, installing and leaning a little about Micro-Cap added a few days.  Actually changing the code to support Touchstone and adding the state machine to perform the narrow sweep was a few hours.  It's hardly what I would consider a major effort compared with some of my other home projects. 

I will be the first to admit that I am not a "real software developer".    I suspect these experts you mention would have it done in half the time with twice the features.   I'm certainly not impeding anyone's efforts and if anything, have offered a few details about my software.   

There is nothing to be gained by keeping the source closed, it will just render your project obsolete as soon as you stop working on it.

I would say the opposite is true in that I have nothing to gain by releasing it.  Maybe you think the fame and recognition of having a few people use it is important to me.  It's not.   The gain in not releasing it is the time I don't spend supporting it.   

To me releasing a .exe is the same as not releasing anything at all.

I doubt I would have made this last demo had Micro Cap not released their simulator.  I could care less if they released the source or not.

I suggest you personally step up and show the community how it's done.  Lead by example. 

[bitseeker]

That narrow-band sampling feature was very cool, Joe. Thanks for adding it to your demo. Even though it takes time to run, it certainly adds to the potential value of the NanoVNA platform. Hopefully the Nano developer can enhance the included software like that, too.

[joeqsmith]

That narrow-band sampling feature was very cool, Joe. Thanks for adding it to your demo. Even though it takes time to run, it certainly adds to the potential value of the NanoVNA platform. Hopefully the Nano developer can enhance the included software like that, too.

I couldn't come up with an better solution.   The latest software has a quasi log sweep where it takes the start, stop and samples per decade, then calculates a series of start and stop segments.  The Nano still performs a linear sweep over each segment but the overall effect is a log.

Without the transfer relay, and me being far too lazy to flip the part, I just fudged S12 and S22 to get the Touchstone working.  Shown is a 500MHz lowpass being swept on the Nano.  The Touchstone is then read by AppCad.  They don't seem to have a dual vertical axis option and I couldn't seem to get it to plot in a linear format but it's obviously working.    Also shown is the 2-port file imported to Micro-Cap and running an AC analysis. 

So, it seems the basics all work. 

I may have some old, low frequency RF transistors floating around.  I could try and find the manufactures S-parameters for them and see if I could come up with the similar numbers with the Nano.  Build and amp and try to simulate that.  A bit more complex but something a little more real world.     

[joedarock]

Joe

Just discovered your work here....very nice! Maybe this question has been asked and answered: Are you going to release it into  the wild so others can use it?

Keep up the great work!

Joe
W3JDR