NanoVNA Custom Software

[realfran]

One other thing I noticed was the sweep rate slow from 10Khz to 400Khz passing this limit start fast.
My observation is with the V.2.8

[joeqsmith]

One other thing I noticed was the sweep rate slow from 10Khz to 400Khz passing this limit start fast.
My observation is with the V.2.8

Sweep rates are dictated by the firmware, not the software and it's really up to Dislord to decide how he wants it to work. 
From the last post, I assume this is what he plans to release:
https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg4024105/#msg4024105

I can let the firmware try and squeeze every drop out while I wait or I can use the original NanoVNA to collect the data in a few minutes and have even higher performance.   
https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg4026451/#msg4026451

[jspencerg]

Hello,
I suggest you remove the V2plus4 from your program header.  Do something to emphasize to new program user that you are in process of tailoring program to the LiteVNA. Previous recommendations of the V2Plus4 should be tempered with your current intentions.  You don't owe users anything, but fair warning of future program changes makes some sense.
I'm getting perspective about the nanovna  ecosystem as I read through this thread.

[joeqsmith]

Hello,
I suggest you remove the V2plus4 from your program header.  Do something to emphasize to new program user that you are in process of tailoring program to the LiteVNA. Previous recommendations of the V2Plus4 should be tempered with your current intentions.  You don't owe users anything, but fair warning of future program changes makes some sense.
I'm getting perspective about the nanovna  ecosystem as I read through this thread.

The header indicates which hardware is supported.   For now, this will be the V2Plus4 and the Lite.  It's not really tailored to one or the other.  It just happens that that V2Plus4 is falling far behind in features.  It sounds like from Dislord's comment that could be addressed easy enough if the developers decided to open it up.  If they do, it's up to the firmware developers to follow Dislords efforts if they want my software to support them.   The same holds true for any future products.  It's up to the developers to make their firmware compatible with the software, if they want to leverage it.     

Because 3.x is so different from previous versions I have created a new repository for it.  I have no plans to do anything more with the original repository.   The 32-bit code will remain for those wanting to run on older OSs.     

[jspencerg]

I thought I'd seen you running current program version with the original nanovna...
So, version 3 will not support the v2plus4? I guess I'll stop reading about the new features with a sense of anticipation.
Thanks for the details.  I realize it's all in the works. I'll stop asking for predictions

[joeqsmith]

I thought I'd seen you running current program version with the original nanovna...

I haven't made any updated to the software for the original NanoVNA since the last release.     

So, version 3 will not support the v2plus4? I guess I'll stop reading about the new features with a sense of anticipation.
Thanks for the details.  I realize it's all in the works. I'll stop asking for predictions

Getting an early start?   

The header indicates which hardware is supported.   For now, this will be the V2Plus4 and the Lite.

[realfran]

For me, the V2plus4 with the close firmware is a great disappointment no possibility to upgrade it, the Lite with the V.3 software is very interesting, especially for use with PC, I need to find some information for the calibration kit, the way to put Kurt calibration Kit data on the software. To calibrate one instrument need one calibration standard the professional one is very expensive out of my possibility. I hope to find someone nearby in the UK with professional VNA to help me to characterize one kit.

[joeqsmith]

Dislord,

Any idea what was changed in LiteVNA64-220225.BIN?   If this contains your most recent changes, can you provide where you ended up with the filter settings. 

***
Last I knew, this was your proposed settings with E0 & E1 defaulting to 15 decimal:

< 20kHz
RBW=6k/(E0 * AVG)

20k - 400k
RBW=12k/(E0 * AVG)

400k - 100M
RBW=60k/(E0 * AVG)

> 100M
RBW=60k/(E1 * AVG)

[joeqsmith]

For me, the V2plus4 with the close firmware is a great disappointment no possibility to upgrade it, the Lite with the V.3 software is very interesting, especially for use with PC, I need to find some information for the calibration kit, the way to put Kurt calibration Kit data on the software. To calibrate one instrument need one calibration standard the professional one is very expensive out of my possibility. I hope to find someone nearby in the UK with professional VNA to help me to characterize one kit.

I have heard of the low cost Kirkby kits but not Kurt.   
https://www.kirkbymicrowave.co.uk/Sales-and-Services/Vector-network-analyzer-calibration-kits/SMA-calibration-kit/

As long as they provide the coefficients with the kit, you should be able to just type them in and hit save.   

[DiSlord]

Yes this firmware contain all changes as in provided here beta.

Only one 'bug' i fix it today, not stable write on some SD cards (i send fixes to Hugen)

[tautech]

For me, the V2plus4 with the close firmware is a great disappointment no possibility to upgrade it, the Lite with the V.3 software is very interesting, especially for use with PC, I need to find some information for the calibration kit, the way to put Kurt calibration Kit data on the software. To calibrate one instrument need one calibration standard the professional one is very expensive out of my possibility. I hope to find someone nearby in the UK with professional VNA to help me to characterize one kit.

https://www.eevblog.com/forum/rf-microwave/diy-short-open-and-load-for-vna-calibration/

[nctnico]

For me, the V2plus4 with the close firmware is a great disappointment no possibility to upgrade it, the Lite with the V.3 software is very interesting, especially for use with PC, I need to find some information for the calibration kit, the way to put Kurt calibration Kit data on the software. To calibrate one instrument need one calibration standard the professional one is very expensive out of my possibility. I hope to find someone nearby in the UK with professional VNA to help me to characterize one kit.

https://www.eevblog.com/forum/rf-microwave/diy-short-open-and-load-for-vna-calibration/

IMHO you'd be better off buying an SMA calibrator kit from Ebay instead of making something. For starters the short is right into the connector so it doesn't add any length.

Random hit on Ebay:
https://www.ebay.com/itm/284154775104

[tautech]

For me, the V2plus4 with the close firmware is a great disappointment no possibility to upgrade it, the Lite with the V.3 software is very interesting, especially for use with PC, I need to find some information for the calibration kit, the way to put Kurt calibration Kit data on the software. To calibrate one instrument need one calibration standard the professional one is very expensive out of my possibility. I hope to find someone nearby in the UK with professional VNA to help me to characterize one kit.

https://www.eevblog.com/forum/rf-microwave/diy-short-open-and-load-for-vna-calibration/

IMHO you'd be better off buying an SMA calibrator kit from Ebay instead of making something. For starters the short is right into the connector so it doesn't add any length.

Yes well until you've attempted to make some and compare their results against a commercially available kit how can one know.
In the thread I've linked there is a wealth of info including some good videos that align with my findings on this matter. SMA fittings are dirt cheap and coupled with SMD components and with a little care it's not hard to make Cal kit components far better that those I've checked that come supplied with any of these cheap VNA's.

[cdev]

I found that making some other fixtures especially -a 1:1 RF transformer is useful. Still have not found a pair of very small alligator or alligator-like strong CLIPS that I am looking for, though. The transformer is to decouple whatever (practically anything) from the transmission line.

Ive never heard of most of those other VNAs. I can see how the nanovna2 is a hard act to follow.

Wonder what it takes to make a pocket sized ultrasound device like I see being sold now.

[realfran]

A DIY calibration kit is one waste of time for two reasons first need one professional VNA with a calibration kit to compare, second many people get some results below 3Ghz.
I think is better to find some good quality of 50homs loads and some clone of short and open (the one that came with the LiteVNA) and play with them, on the end the low-cost VNA got is on limit.
I believe one good software can better use the resources of them using PC power computation.

[joeqsmith]

A DIY calibration kit is one waste of time for two reasons first need one professional VNA with a calibration kit to compare, second many people get some results below 3Ghz.
I think is better to find some good quality of 50homs loads and some clone of short and open (the one that came with the LiteVNA) and play with them, on the end the low-cost VNA got is on limit.
I believe one good software can better use the resources of them using PC power computation.

Your previous referring to Kurt must have been someone who characterized a set of the included standards and not a cal kit they sell.  If that's true, I suggest you contact them and see if they would be willing to supply you with the coefficients. 

Your point about needing to compare the home made set with a known setup is valid.  However, I just want to remind you that everything I have shown with the low cost VNAs has been with the included shorts, opens and load(or a hand selected load).   Or, in many cases when testing below a GHz and looking at components, a set of standards I made up on FR4 were used.   In all cases I have used the ideal model.   

I could certainly attempt to characterize these standards and it may even improve my measurements some measurable amount.  However, in your case I suspect there are much bigger problems that would mask any gains.   You posted some data for an attenuator that showed several dB of error.  No cal standard can correct that problem. 

Besides myself, someone else had posted a report for you making a similar measurement with their low cost VNA.
https://nanorfe.com/forum/attachments/Calibration/96e7e95e1885-0/VNA_CalCheck_Summary_12-5-2021.pdf 

I would expect you to get similar results.  I've suggested you simplify your setup by removing all your home made bits until you start getting better data but I suspect you think that the calibration is going to fix it all for you.   That book I mentioned earlier is a pretty good starting point and covers a lot of basics.   

https://www.google.com/books/edition/The_VNA_Applications_Handbook/_VyzDwAAQBAJ?hl=en&gbpv=1&printsec=frontcover

 

[realfran]

Yes, Kurt Pulsen has characterized a set for the liteVNA- LibraVNA and  VNWA  and many others.
The name of the software is "NanoVNA-Solver64" is good have it is a proper name 
For the calibration, I use to do it first directly on the LiteVNA and after thru the HP switching test set, watching for the different results with and without the switching relay,
the isolation is 100dB and even there is some problem the LiteVNA does not see them.

[joeqsmith]

We kicked around a few names but Solver64 won out as it best described it.  Unlike all the previous software I have released for these low cost VNAs, version 3 is a 64-bit application.  This means a new runtime engine and VISA.

Its possible you are fighting a cable, connector, relay, attenuator ... problem.  There's just no way that an attenutor would be off by as much as you show without there being a major problem.   

With the release of Dislord's firmware, I have started to work on a demonstration video for the new software.  Because of Dislords added features, the Lite will be used during the demonstration.  There's a lot to go over.  I plan to revisit the transfer relay and T-Check.   Just for you, I'll show a step attenuator. 

The data that OWO had presented for the V3 showed 100dB.  They were planning to push it further. 

[realfran]

The story of OWO OWO on the NanoVNA V3 takes time, I don't like the close source FW like the NanoVNA 2 Plus4 and the hysterical version Plus5 with some improvement (on the firmware) to scrub some money with this supposed version $329.00 and now quick disappear and recall Plus4 Pro from the Nanoforte web page.
My question is: what is the reason for Plus5 improvement? get more money? instead to upgrade Plus4  with new firmware.
One other VNA 6Ghz with (supposed 100dB) with squarewave generator for  $1300 [full two-port version] on 5-12 mont time is making noise and loss of credibility.
I hope Dislord can get another 10dB more from the LiteVNA and for the hobby, use is good.

[joeqsmith]

Most of my test equipment is closed source.   It doesn't concern me as long as it does what the manufacture claims.  In the case of the V2Plus4, the mail group is not public.  The last time I posted I was told to not to bring up the Lite.  I then made the choice to leave the group and have no plans to return.   We were exchanging PMs for a while but similar to what you saw her posting here a few pages back about business practices.     

I assume any successful business is about monetary gains.  If they achieve 120dB that's impressive.  They claimed to support the unknown thru.  I doubt the hardware architecture is at all the same.   If they are able to get reasonable performance do to 10kHz and support narrow band measurements, I would buy one.  Even with the squarewave drive. 

The Lite is very slow and can't out perform even the original NanoVNA in the lower frequencies.  Then again, I assume most of their market are CB and amateur radio operators.   I once had an electrical worker ask me why would I ever need to measure frequencies above 60Hz.    The CB/hams perspective seems to be that the VNA is an antenna analyzer.   We all have different perspectives.  OWO talked about targeting smaller companies, not the radio hobbyist. 

[joeqsmith]

Flipper's second Lite arrived.   Sadly, same problem as his original.   Sounds like the supplied USB cable was also bad.   Measuring the voltage at the (HP Z640) USB connector at 5.098V.   There is also a rapid charge port which outputs 5.25V.  Both are too high for the LiteVNA to work properly but note that there is a difference. 

Looks like the diode dropped powered hub is the fix for now.  Flipper is sending me the VNA and I will attempt to make the Changes Dislord mentioned and see if we can improve the low frequency performance.

[realfran]

LiteVNA64: Remove D1 R7 R1 -Please disconnect the battery and USB power connection when modifying the circuit. The battery must be connected after the modification is completed to work properly.
This is the modification suggested.
Another suggestion is: Calibrate without a USB cable (the reason is on calibration is embedded in the noise).
I calibrate the LiteVNA on battery with USB cable disconnected I started the NanoVNA-Solver V2.8 and I get this.

[joeqsmith]

I was referring to the front end modifications Dislord had mentioned:

https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg4015222/#msg4015222

In Lite if remove DC block capacitor and ESD protection diode possible get better results on low frequency range < 10M (but this huge increase chance burn rf switch)

I have no plans to modify the power supply section.   I had asked a few questions to try and understand if these changes would cause other problems but we did not get a response.   So for now, Flipper will stay with the low voltage powered hub.   

https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg4030855/#msg4030855

***

Another suggestion is: Calibrate without a USB cable (the reason is on calibration is embedded in the noise).
I calibrate the LiteVNA on battery with USB cable disconnected I started the NanoVNA-Solver V2.8 and I get this.

The data I've presented in regards to the power supply noise has all been without calibration.  Calibration isn't a panacea able to overcome such hardware problems.   

That said,  we always run them with a PC.  We expect raw data from the VNA and calibration is handled by the PC.  It should be easy enough to replicate. 

***
Snip a USB cable and expose the power leads.  Power the VNA from an external supply.   Flipper's will start to act up with 5.04V applied.  I don't know the upper voltage limit before damage but it was plugged into that charge port.  Keep it below 5.2V and I suspect you are fine.   
Assuming no calibration, I would expect you to measure 50-60dB in the 400kHz to 1MHz range. 

Then you can try to calibrate your way out of it.  Let me know how it works out for you. 

[realfran]

On the PC software, I don't perform any calibration of their raw data from the VNA

[joeqsmith]

On the PC software, I don't perform any calibration of their raw data from the VNA

I understood that from the photo and what you had wrote.

Another suggestion is: Calibrate without a USB cable (the reason is on calibration is embedded in the noise).
I calibrate the LiteVNA on battery with USB cable disconnected I started the NanoVNA-Solver V2.8 and I get this.

However, you made the suggestion which suggests to me that you do not understand some of the basics.   

Run through a calibration on battery (stand alone, no PC), but do not insert any of the standards when doing it.  Create a bad calibration.  Save it to 0.     Reset the unit and it should show a lot of random garbage.  Now connect it to the PC and run Solver.   If you see the same random garbage, then something is wrong!!   Again: 

We expect raw data from the VNA and calibration is handled by the PC.

I had discussed this with Dislord a few pages back.  There was a time where he would not send the raw data but this was corrected: 
https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg3947854/#msg3947854

The data that you show is what I expect.