NanoVNA Custom Software

[Alextsu]

Hi, Joe,
There is now an alternative FW for all V2-2 versions, ported from NanoVNA by Dislord,
check this link

https://groups.io/g/nanovna-users/topic/84427595

Though haven't tried it yet

[joeqsmith]

Hi, Joe,
There is now an alternative FW for all V2-2 versions, ported from NanoVNA by Dislord,
check this link

https://groups.io/g/nanovna-users/topic/84427595

Though haven't tried it yet

I had spent a fair bit of time trying some of their firmware on an H4 I had purchased recently.  I would have no reason to try their firmware on other devices.   If you are interested in reading more about that testing:

https://www.eevblog.com/forum/rf-microwave/nanovna-custom-software/msg3579600/#msg3579600

The V2Plus has ran for 24 hours using various settings without a single lockup.     It's never been this stable.   Next step is to see if it will put out some good numbers. 

If you don't have access to the groups.io because of their dick move to lock it, use the mirrored site:
https://nanorfe.com/forum/V2-Plus-Firmware.html

[JohnG]

Thanks for the software and the excellent documentation. I have had no trouble following the instructions and getting the software up and running with my NanoVNA V2Plus4.

I do have one difficulty, which has to do with readability. On my Win 10 computer with 1600 x 1200 monitor, the software window is small and hard to read. After playing with many settings,  the only workaround I can find is to change the monitor resolution to 1280 x 800, which helps some, but messes up my Windows desktop (rearranges the icons), and adds a little blur to the display.

Do you know if this is a fundamental limitation of the National Instruments platform used to develop the software? If so, I can at least quit fiddling with it.

Thanks,
John

[joeqsmith]

Thanks for the software and the excellent documentation. I have had no trouble following the instructions and getting the software up and running with my NanoVNA V2Plus4.

I do have one difficulty, which has to do with readability. On my Win 10 computer with 1600 x 1200 monitor, the software window is small and hard to read. After playing with many settings,  the only workaround I can find is to change the monitor resolution to 1280 x 800, which helps some, but messes up my Windows desktop (rearranges the icons), and adds a little blur to the display.

Do you know if this is a fundamental limitation of the National Instruments platform used to develop the software? If so, I can at least quit fiddling with it.

Thanks,
John

Who knows what you were expecting.  Maybe you expect the graphs to scale.  Obviously the font sizes are fixed.   There are videos on YT about designing software with LabView using the autoscaling if you are interested in learning about it.   

I've run my VNA software on 1280X1024 19", 1920X1080 27" and 1920X1200 24" and its very easy for me to read on them all.   It is designed to run on my laptop which has a 17" display.   

You should consider running  one of the popular open source programs.  If you really have some need to run my software and your choice of monitors was poor for your eyesight,  just buy a new monitor.  They are cheap.     

[neilhao]

I just gave this firmware a shot. Generally, it is amazing with more features, however it still has bugs for now.
I noticed this firmware could not measure the correct results in low freq band on the V2_2 hardware, e.g. from 50kHz to 100kHz. The official firmware has no problem.

Hi, Joe,
There is now an alternative FW for all V2-2 versions, ported from NanoVNA by Dislord,
check this link

https://groups.io/g/nanovna-users/topic/84427595

Though haven't tried it yet

[DiSlord]

> I noticed this firmware could not measure the correct results in low freq band on the V2_2 hardware, e.g. from 50kHz to 100kHz. The official firmware has no problem.
This can related to output power for SI5351 (i add option for select it, See STIMULUS->CFG SWEEP->TX POWER, for low freq range need select minimum SI -10dB), i add fix for this in next.

PS also in this fw real frequency step in < 140 MHz range = 1Hz

[JohnG]

Who knows what you were expecting.  Maybe you expect the graphs to scale.  Obviously the font sizes are fixed.   There are videos on YT about designing software with LabView using the autoscaling if you are interested in learning about it.   

I've run my VNA software on 1280X1024 19", 1920X1080 27" and 1920X1200 24" and its very easy for me to read on them all.   It is designed to run on my laptop which has a 17" display.   

You should consider running  one of the popular open source programs.  If you really have some need to run my software and your choice of monitors was poor for your eyesight,  just buy a new monitor.  They are cheap.   

It appears the display resolution of the software interface is fixed. That's the answer I was looking for, and I have no further expectations because it is free.

John

[joeqsmith]

Who knows what you were expecting.  Maybe you expect the graphs to scale.  Obviously the font sizes are fixed.   There are videos on YT about designing software with LabView using the autoscaling if you are interested in learning about it.   

I've run my VNA software on 1280X1024 19", 1920X1080 27" and 1920X1200 24" and its very easy for me to read on them all.   It is designed to run on my laptop which has a 17" display.   

You should consider running  one of the popular open source programs.  If you really have some need to run my software and your choice of monitors was poor for your eyesight,  just buy a new monitor.  They are cheap.   

It appears the display resolution of the software interface is fixed. That's the answer I was looking for, and I have no further expectations because it is free.

John

Again, I'm not sure how you expected it to work.  Most software I use will not scale the fonts.  It could be done  (selecting a different font for example) but it's uncommon.  LabView will allow most graphics to be scaled.   Currently the software uses graphs that were manually drawn using a CAD program and saved off as raster images.   If I settled on the standard graphics, these could easily be scaled but I wanted my own graphics.   

Buying a monitor that is easy for you to read would seem like a must, just for your normal use.

****
Windows 10 does have that magnifier built into it.  Really a poor solution IMO, but it's there.   

[joeqsmith]

I finally had some time to put a short video together showing flippers printed torque wrenches.   

[ALW]

Thanks a lot JoeSmith for this excellent SW. I am learning to use it... Available options are massive, but I believe it is very powerful.

As for the readability issue on high density screens, I think I found a solution in Win10 that works nicely.
In C:\Program Files (x86)\NanoVNA_V2Plus right-click the .exe, select "properties" tab "compatibility"
Click "Change high DPI settings"
Select "High DPI scaling override" option "System (enhanced)"

In my case that made the display crisp and clear and very well readable. Only drawback is that some vertically rotated graph legends were cut off, but I can imagine what they say.
Best
Al

[joeqsmith]

As for the readability issue on high density screens, I think I found a solution in Win10 that works nicely.

John was having problems with a 1600 x 1200 monitor.   I have used it on monitors up to  1920X1200 24" with no problems.   John fails to provide details about the size the monitor they are using.   If my crystal ball would ever get repaired then maybe I could see it was an 4" LCD.     

I would expect with the display and keyboard being most peoples primary interface to the PC, that you buy something that fits you.  Especially if you use the computer often.   There are 1600X1200 17" monitors fairly cheap but with my ageing eyesight, at that ratio I would have no use for one.   

[ALW]

Forgot to mention that I am using a Surface with 2160x1440 resolution. The unmodified / standard magnified window was blurry and difficult to read. With the settings above it is crisp and clean.
BR

[joeqsmith]

Forgot to mention that I am using a Surface with 2160x1440 resolution. The unmodified / standard magnified window was blurry and difficult to read. With the settings above it is crisp and clean.
BR

12" display?   Do you have the docking station with a separate monitor?

[ALW]

10" display, no docking

[joeqsmith]

Wow!!  No wonder you were having problems viewing it.   Maybe John was attempting to run it on a small tablet as well.   

[ALW]

@joeqsmith thanks again for sharing this incredible piece of SW. I still can't believe I am using a little cheap box with this professional interface.
I am trying to analyze the recorded sweep data in Igorpro for prototyping a medical radar application (trying to visualize heartbeat / breathing by analyzing the phase of the signal reflected by the body surface).
If you find the time, could you provide some short hint how the binary sweep data are structured in the recorded sweep file? Little / big endian, byte index, port1 port2 interleave,...
Thanks a lot in advance, best regards
Albrecht

[joeqsmith]

Someone else asked a similar question.   If you know what format you need, document it.  Make sure you include every detail.  It can be ASCII or anything else you want.     

[ALW]

What I need would be similar to a touchstone s2p file:
Columns
Freq Re{S11}  Im{S11}  Re{S21}  Im{S21}
or alternatively
Freq MagS11 PhS11 MagS21 PhS21

Lines
Freq1
Freq2
...
Freq1
Freq2
...
for continuous sweeps

If it is ASCII or binary doesn't matter, I could read both as long as I know the header / offset and byte structure of the binary.

[joeqsmith]

Which software are you are using?  I may add it to the main program is why I ask. 

Let's assume ASCII, do you want the frequency in Hz?  SI units?  What?    Same for Mag in dB and phase in degrees? 

Are you wanting to use a SPACE, 0x20 for the delimiter as shown? 

How do you want to mark the end of line?  CR?  CRLF?  LF?


***
Also, I don't understand the Freq1 showing up twice in your example.

[ALW]

First, thanks a lot for your help! I really appreciate it.
I am using a software called IgorPro, that I am familiar with. It is similar to Mathlab, but is more powerful in doing complex math in (almost) real time, therefore cool for prototyping, later I am planning to retrieve the data directly using the com port of the Nanovna. Igor can read any data structure as long as I can tell it the structure.
So for your questions, anything that is most convenient for you when programming. Freq in Hz Mag in dB and phase in ° would be fine. It is however no problem to convert from other units if this would save your time. Space as delimiter and CR EOL would be fine too. Or if binary the byte structure of the float and the indexing / header length would be all that is required to read it.
With the freq1 showing twice (looping) I mean the recording should loop through the frequencies for multiple repeated sweeps... Of course it would be enough to have the frequencies only once, they do not need to be repeated

[joeqsmith]

When I asked

Which software are you are using?  I may add it to the main program is why I ask. 

  I am referring to my software.  I would have no way to add anything to IgorPro.   No matter as it sounds like you shouldn't have any problems then working with the binary file directly. 

FREQ (double)
CHANNEL0&1 (double complex)

The data is contiguous.

[ALW]

Sorry, I misunderstood you. Using your latest release r2.0.3 with a NanoV2+4.
Should be able to read your binary, but do not understand yet. Your bin file seems to have a header, do you know how long that is / where the data block starts?
FREQ (double)
CHANNEL0&1 (double complex)
means
FREQ (double)
CHANNEL0(double complex)
CHANNEL1(double complex)
?
I am reading the Labview specs right now, I think I am finding the definitions. Looks like double complex is 64bit real 64bit imaginary. Will try to figure it out, thanks a lot!

edit: With the freqs being repeated I should be able to find the data block/ header end. Will bug you only again if I can't figure it out

[joeqsmith]

Yes, channe0 followed by channel 1.  Yes, double is 64-bit.  Yes complex will store both the real and imaginary component to two separate 64-bit numbers.    The header is from prepending the array size.  Data is stored big-endian.

[ALW]

Thanks a lot. That will do

[joeqsmith]

If you get stuck, let me know.  It would only take a few minutes to write a translator.   

***
Good to see someone attempting to use it for something other than VSWR.    I was trying to measure various liquids with the V2Plus4 using a simple home made coaxial probe. 

https://www.mdpi.com/2075-4418/8/2/40/pdf

https://www.keysight.com/us/en/product/N1501A/dielectric-probe-kit.html