NanoVNA Custom Software

[joeqsmith]

Sweeping from 18 to 28C, from the paper I previously attached we would expect to see less than 0.1db and 1degree of change.  While we could cherry pick some frequencies, it seems both are very stable with temperature.

[neilhao]

trust me when you come up with your V3 with -200dB noise floor at $50 but doesnt support NVNA-QT or doesnt have built it characterized calibration kit setting, i'll find that less interesting or less usefull atm.

I still doubt about the performance of V3. 500 USD is more than reasonable if V3 has thus kind of performance.
The only problem is when I could place the order. No ETA for V3 for a long long period. Thus, I just placed an LiteVNA order for digging in its RF design, maybe its Noise Figure could be improved by user community...Refining the PCB and SMT becomes far easier than before.

[Mechatrommer]

trust me when you come up with your V3 with -200dB noise floor at $50 but doesnt support NVNA-QT or doesnt have built it characterized calibration kit setting, i'll find that less interesting or less useful atm.

I still doubt about the performance of V3. 500 USD is more than reasonable if V3 has thus kind of performance.

that is just my too optimistic exaggerated guess but in real life impossible to achieve even by #1 brand... for eq Keysight FieldFox is -100dB dynamic, Siglent is -120dB iirc... i've been shown the result of prototype board thats on par or very close to those but i'm not allowed to post it in public. cheers.

[joeqsmith]

OWO shared some preliminary data from the prototype with me as well.  We discussed making a review comparing the V3 prototype with the LiteVNA to help stir up some interest in the product.  Once she made the comment: 

Or I may just go and take the 6 figure job at a big technology company doing cutting edge R&D and be gone with this race to the bottom crap, maybe I'll decide that's where I want to be in life instead.

, I felt the risk was too high that they wouldn't follow through and it made no sense to continue down that path.   Her constant lecturing me of business practices further cemented my concerns.  I hope I am wrong.   

I have added support for setting the output power to support the LiteVNA and also increased the timeout to handle the very slow data transfers when it is set to low frequencies.   The power settings will be stored into the defaults.  This means that both the defaults and calibration files are not compatible with the 2.x software. 

[OwO]

Alright, thanks for sharing your points of view.

About V3, as I said it's a hobby project and please do not take it too seriously. I have no intention of trying to compete with them for market share and I don't necessarily want to "stir up interest". I didn't start that thread and only ended up posting in it because too many people were emailing me and giving me time pressure that I really do not need. V3 will happen when it happens, and in the mean time look at other alternatives too. EuroVNA is another one possibly worth following.

[joeqsmith]

0xF0 - (byte) device variant (2 vor V2)
0xF1 - (byte) protocol version (1 vor V2)
0xF2 - (byte) board revision (2 vor V2, 3 - V2Plus, 4 - V2Plus4)
0xF3 - (byte) firmware major
0xF4 - (byte) firmware minor

Was there a reason for not using a unique board revision for the LiteVNA?   
Or do you consider the 2, the lite? 

[DiSlord]

>Was there a reason for not using a unique board revision for the LiteVNA?   
I leave this values
#define BOARD_NAME "LiteVNA 6G"
#define BOARD_REVISION (2)
#define FIRMWARE_MAJOR_VERSION 2 <--- Change only this for Lite (V2/V2Plus/V2Plus4 = 1)
#define FIRMWARE_MINOR_VERSION 2

For compatibility with programs on the CPU side (i tried to change everything about the protocol to a minimum). Software like NanoVNA-App detect device type from this registers.
PS LiteVNA as V2Plus4 can provide up to 65535 points in one sweep

For my fw version i use LiteVNA code for compile V2 fw, this allow me get more measure and calibration points (up to 401) and all features from H / H4 devices (also if need V2 can modded for add SD card support)

[joeqsmith]

>Was there a reason for not using a unique board revision for the LiteVNA?   
I leave this values
#define BOARD_NAME "LiteVNA 6G"
#define BOARD_REVISION (2)
#define FIRMWARE_MAJOR_VERSION 2 <--- Change only this for Lite (V2/V2Plus/V2Plus4 = 1)
#define FIRMWARE_MINOR_VERSION 2

For compatibility with programs on the CPU side (i tried to change everything about the protocol to a minimum). Software like NanoVNA-App detect device type from this registers.
PS LiteVNA as V2Plus4 can provide up to 65535 points in one sweep

For my fw version i use LiteVNA code for compile V2 fw, this allow me get more measure and calibration points (up to 401) and all features from H / H4 devices (also if need V2 can modded for add SD card support)

Other firmware could end up with the same Major Version as the Lite.   Device variant or board revision may be a better choice to make a unique ID for the Lite.   
https://nanorfe.com/nanovna-v2.html
They advertise 1024 points with the external software.   Maybe you are suggesting with your firmware that the V2Plus4 could be increased?   

[DiSlord]

>Maybe you are suggesting with your firmware that the V2Plus4 could be increased?   
Strange, i remember for V2Plus4 on this site write 65535 (original fw also V2Plus4 allow send this points count)

Question not owerflow input buffer:
Example for get N points need write in 8 bit register N (8 bit limit 255)
So for read 1024 points need write 255 again 255 again 255 again 255 and 4
For 65535 points need lot of writes (and possible owerflow)

I allow write 0 in this register one time and then Nano return all sweep points (as i remember this fix also present in official firmware)

For LiteVNA and V2Plus4 possible set up to 65535 points

PS my fw variant for V2Plus4 not work on new V2Plus4 revisions (used different LCD module? i dont`t know details, only know CPU connection work)

[joeqsmith]

>Maybe you are suggesting with your firmware that the V2Plus4 could be increased?   
Strange, i remember for V2Plus4 on this site write 65535 (original fw also V2Plus4 allow send this points count)

I'm just going by their document.  Today I limit my software to 1600 points but it's one number change to open it up.   
 

Question not owerflow input buffer:
Example for get N points need write in 8 bit register N (8 bit limit 255)
So for read 1024 points need write 255 again 255 again 255 again 255 and 4
For 65535 points need lot of writes (and possible owerflow)

Sorry you lost me.  I have no idea what 8-bit register you are referring to. 

I allow write 0 in this register one time and then Nano return all sweep points (as i remember this fix also present in official firmware)

Sorry, again I have no idea what register you are referring to.   

From:  https://nanorfe.com/nanovna-v2-user-manual.html  , Sweep points is a 16-bit register.  I just set that to what ever I need and the device returns that number.     

For LiteVNA and V2Plus4 possible set up to 65535 points

PS my fw variant for V2Plus4 not work on new V2Plus4 revisions (used different LCD module? i dont`t know details, only know CPU connection work)

Yes, its a 16-bit register but from their specs, I am not sure what they actually support.  I can make a quick change to my software and try to see what the limits are of the V2Plus4 (with what ever firmware I have installed).   I've never upgraded it.

[DiSlord]

> Sorry you lost me.  I have no idea what 8-bit register you are referring to.
https://nanorfe.com/nanovna-v2-user-manual.html
See: Host to device command list
READFIFO - for read data from V2
18 (AA) (NN)
How you ask V2 for read data? (NN) - 8 bit value

[joeqsmith]

I set it to read one value at a time (32 bytes).   

Both seem to support 16-bits.  V2Plus4 is much slower.  However, I can't think of a need for it off hand.   I may go ahead and open it up on the next release. 

[DiSlord]

>I set it to read one value at a time (32 bytes). 
You can set up to 255 values (can be faster)
Or set 0 then get all sweep points

>V2Plus4 is much slower.
Possible this related to your read procedure? 65535 points must read in near 3 min on default avg setting (400 points/sec for V2Plus4 and 550 points/sec for Lite).
You can try ask 255 data point in one time (or try use 0 value for get all sweep points data, last firmwares for V2Plus4 should support this)

Example read 1020 points (by 255 points in one ask):
Send: 0x18 0x30 0xFF 0x18 0x30 0xFF 0x18 0x30 0xFF 0x18 0x30 0xFF
And wait 1020 data points

Can try this
Send: 0x18 0x30 0x00
And wait all sweep points

[joeqsmith]

>I set it to read one value at a time (32 bytes). 
You can set up to 255 values (can be faster)
Or set 0 then get all sweep points

>V2Plus4 is much slower.
Possible this related to your read procedure? 65535 points must read in near 3 min on default avg setting (400 points/sec for V2Plus4 and 550 points/sec for Lite).
You can try ask 255 data point in one time (or try use 0 value for get all sweep points data, last firmwares for V2Plus4 should support this)

Example read 1020 points (by 255 points in one ask):
Send: 0x18 0x30 0xFF 0x18 0x30 0xFF 0x18 0x30 0xFF 0x18 0x30 0xFF
And wait 1020 data points

Can try this
Send: 0x18 0x30 0x00
And wait all sweep points

For the V2Plus4, the sweep time is what dominates, not the USB communications.  I've ran several tests collecting large data sets and the software easily will keep up, even asking for a single value.

Yes, they are very close to the 400Hz, or 2.5ms per value.   Of course, you need to be using the higher frequency range.  Looking at the attached, we are sweeping from 500MHz to 900MHz with 4000 points.   The software is measuring 10 seconds per sweep as expected. 

I wouldn't mind being able to pull down 4000 points in a second but again, I can't think of a case where I needed this much data from a linear sweep.   

***
Some time back, a member was attempting to use the V2Plus4 to measure human vitals which required running CW at a fixed sweep rate.  They were able to get it working.  It's a few pages back if you are interested in having a look. 

[joeqsmith]

Based on 

550 points/sec for Lite

,  4000points would require 7.3 seconds.    Oddly, it is very slow.    I tried it at 400 data points and it is about 10X faster but no where near what I would expect.   The V2Plus4 is much faster.

[DiSlord]

LiteVNA use 2x avg as default, this give ~300 points/sec

You can control this by write avg value in 0x40 register:
Set average = 5x command:
0x20 - write byte
0x40 - address (average)
0x05 - value
Need send : 0x20 0x40 0x05

High avg setting slowdown measure speed, but increase quality. You can also do average inside your software.

[joeqsmith]

LiteVNA use 2x avg as default, this give ~300 points/sec

You can control this by write avg value in 0x40 register:
Set average = 5x command:
0x20 - write byte
0x40 - address (average)
0x05 - value
Need send : 0x20 0x40 0x05

High avg setting slowdown measure speed, but increase quality. You can also do average inside your software.

Doing a running average in software, I maintain the VNAs data rates.  So for 101 points and 10 averages, I would have a 101 X 10 array that maintain. 

It would have been nice to see it backwards compatible with the V2P4 so no software changes were required.  It seems like in this case, it seems like you could have used different defaults.  Odd is until now, you did not mention it and there was nothing in that short document you provided that mentioned the defaults.   I get to stumble into it.   

I'll add the new command and try it.
***
What does your firmware do when I set this to 0?

Does writing to this register instantly reset the sweep and average? 

Set it to say 4000 points, 100 avg.  Switch it back to one, does it hang?  Mine does.  I can disconnect and reconnect and yet can not get the VNA to start sweeping again.  Is this expected by design? 

Does the firmware require that you stop sending data before changing the average (or some other requirement)?  Is it similar to the power selection that requires a sweep rather than CW to update?

[DiSlord]

>It would have been nice to see it backwards compatible with the V2P4 so no software changes were required.
Just use default, result just little less sweep speed 300 points / sec
This just additional option (my fw for V2Plus4 also support this, but not remember about official)

>What does your firmware do when I set this to 0?
I check limits, min value = 1

>Does writing to this register instantly reset the sweep and average?
Yes, sweep reset after last asked data send to CPU.

V2 protocol parse input commands and put in FIFO buffer, then commands executed
So if you ask data Nano start send it, received command put in FIFO and wait, after send complete executed next command. FIFO have limited depth so can overflow (in this case Nano not accept new commans while old executed and free buffer)

PS I check this option work vs NanoVNA-App application, and not see any problems. And repeat - use can use this default settings, it will not affect the work in any way.
PS Also i fix output power change in CW mode (and additional fix update XTAIL correction, this allows you to set the output frequency up to 140 MHz with an accuracy of 1-5 Hz), need some tests.
Change power not apply in CW mode - i cache generator setting, and not update if no frequency change, as result no power settings update. Now i reset cache after receive this command.

[joeqsmith]

>Does writing to this register instantly reset the sweep and average?
Yes, sweep reset after last asked data send to CPU.

With a 100 averages, that could take a VERY long time.  You need to have a way to abort and recover.   

Setting the average with the firmware included with my V2Plus4 has no effect.

What about the lockup I mentioned?  Can you replicate it?  Seems like I can easy enough.  Maybe it's unique to this firmware. 

I really don't know what the implications are of averageing to everyone else.  Personally, if something isn't stable, I want to see the peak heights.  I want the raw data.

Changing the sample rate, hard to say what the effects are.   

[DiSlord]

I think I already wrote here what averaging is on the nana side. Perhaps the wrong term here is averaging.
This value affects the bandwidth in which the measurement is performed (this is not exactly what you do in your program, but it looks like)

Let’s be more specific, nana digitizes the signal for say 10 periods, this data is sent for processing for calculation. But if you take not 10 periods but 20, then the value will be calculated more accurately. If we consider this as a spectrum analyzer, then the analogy is a decrease in RBW during measurement, this reduces the noise level, but unfortunately more samples need to be taken.

So in any cases you get RAW data

>Can you replicate it?
I can test only in NanoVNA-App.
If i set 1000 points and x100 avg, then run sweep and disconnect - then Nano not hang, after reconnect i can run.

As i see nanoVNA-App reset sweep before disconnect and on sweep start (by send 18 30 00)

Here program log in this case:
10.159 tx: setting scan parameters
10.163 tx: 00 00 00 00 00 00 00 00 23 00 80 D1 F0 08 00 00 00 00 23 10 CB 41 57 00 00 00 00 00 21 20 00 04 21 22 01 00
10.166 tx: requesting 1024 points ..
10.168 tx: 18 30 00                                   <- Here i disconnect
11.966 tx: clearing FIFO buffer
11.969 tx: 00 00 00 00 00 00 00 00
11.972 tx: leave USB data mode
11.975 tx: 20 26 02
11.978 tx: polling - requesting hw-revision and fw-version
11.981 tx: 10 F2 10 F3 10 F4
11.984 disconnect: serial port disconnected
15.305 connect: serial port connecting ..
15.308 connect: serial port connected                <-- reconnect
15.337 tx: leave USB data mode
15.340 tx: 00 00 00 00 00 00 00 00 20 26 02
15.343 tx: are you a V2 ?
15.346 tx: 0D
15.349 rx: 32 .. ACK OK
15.352 tx: requesting device variant
15.354 tx: 10 F0
15.357 rx: 02 .. DEVICE VARIANT OK
15.360 tx: requesting serial protocol version
15.363 tx: 10 F1
15.399 rx: 01 .. PROTOCOL VERSION OK
15.402 tx: requesting hw-revision and fw-version
15.405 tx: 10 F2 10 F3 10 F4
15.460 rx: 02 ..... HARDWARE REVISION
15.463 rx: 02 02 .. FIRMWARE VERSION
15.466 rx: [connected to a V2]

>Maybe it's unique to this firmware.
No i nothing change in fw related to protocol (command and data processing equal as on any V2, just added additional commands).

Test on Current version of your software, i manually set on Lite avg = 80x, run sweep on software, then disconnect (remove cable), connect cable back and reconnect, after i can run sweep as before

But if i run sweep stop it, restart. In some cases sweep not start. Also i see after restart sweep point counter (point index) not start from 0 after sweep resume after stop on 50 for example after resum it can start from 120 (not sended reset sweep?) But if disconnect in software and reconnect, sweep start from 0.

PS But then test USART connection to LiteVNA (connect to Lite over Bluetooth and WiFi), i found bugs - on low speed connection V2 does not have time to send data, and fills its buffer for measurements, as a result, it overflows and data is lost, as it is overwritten with new ones. I fixed this by pausing measurements while filling.

[joeqsmith]

Plotting the sweep time for the two VNAs.  Both set to 401 points, 150-300MHz.  I would imagine the stability and averaging are not very important to the antenna analyzer group.  For our friend attempting to do their radar experiments, maybe its a bit more critical. 

I did try changing the process to realtime but as expected, no effect. 

So I set the average back to 2 and the system is stable.  It could be a problem with my software.  Hard to say. 

Where are you storing the firmware?   I would be willing to have a look at the new version that supports the narrow band measurements. 

[DiSlord]

I made firmware for self (but send last code to Hugen, he additionally checks it on his devices and assembles the firmware)

I don't have the latest hardware revision, just a couple of engineering samples (2.8 inch version and very old 4 inch board)

Official page here:
https://zeenko.tech/litevna

I can build my last code
LiteVNA 62 for 2.8 inch devices (i test it on my sample)
LiteVNA 64 for 4 inch devices (must work but not test)

Fixed update power setting in CW mode for both SI and MAX
Added more trace/marker types
Added USART connection support

>narrow band measurements
If i correct calculate 1xavg give 2kHz bandwidth on > 400kHz frequency (

On < 400kHz used 12k IF on less 20k use 6k IF so measure on this range slow (very slow).
Min frequency = 1.6k but in this range Lite have very bad dynamic (i allow use it only for tests LiteVNA Specifications 50kHz ~ 6.3GHz).

All V2 (and LiteVNA) on ~ < 1Mhz not good.
H4 up to ~100MHz work more better then any V2 variant include LiteVNA

[joeqsmith]

I have the 4 inch version.   No problems installing the firmware.  The only thing I had to do was calibrate the touch screen (so I could check the version).   Measuring a crystal, I get the same values as I did with the previous firmware.   Yes, very slow at lower frequencies.   I'll try out the power level settings.

[DiSlord]

On serial crystal measure i get better results if select x5 or x10 avg

LiteVNA (as H/H4) have crystal measure calculator (see MEASURE->SERIES XTAL)
I use Phase Shift Measurement https://www.mikrocontroller.net/attachment/473317/Crystal_Motional_Parameters.pdf method for calculate
For get better results on limited points count i use bilinear interpolation on search

Here it results for 10MHz serial xtal connection
NanoVNA-H (101 points)

Here some measure on LiteVNA (1001 points allow better search resonances and phase shifts):

V2Plus4 not allow made this measure (it use additional processing for get better dynamic, so results look strange):


But any V2 not good in XTAL measure, due to used measure mode (first measure reference, second refflect, second thru), H/H4 measure (reference / reflect and reference / thru in some time, and this allow see correct XTAL responce)

Lower frequency XTAL better measure on bigger avg settings (results look better)

[joeqsmith]

I've not tried to use any of the built-in features for the NanoVNAs.    It may be of interest to compare the LiteVNA against my original NanoVNA.

What do you require for a test fixture?  I see the paper you provide has a few.  I made a few different fixtures but didn't obtain very good results.  The last fixture I constructed  seemed to get me very close results when compared with other methods:

https://www.eevblog.com/forum/rf-microwave/fun-with-crystal-filters/msg3076877/#msg3076877

I have a couple of crystals that I had collected data for but was never able to get my hands on a reference standard.