开云体育

Somehow my photo of NanoVNA with STM32F303CCT6 and the USB mod uploaded to "files/Hardward Mods" didn't show up there. So I attached my photo below. The PCB was provided by Hugen, since I don't have steady hands to solder such tiny wire and component myself.

A 1.5K reistor was soldered between pin PA12(USB_DP) and VDD so that host PC can recognize the NanoVNA for serial comminication. The C6 decap nearby has a VDD pad near MCU. The PCB is full of flux because Hugen reworked it to replace the MCU to F303, and I did it again to clean up the solder bridge underneath the MCU.

I reaplced the STM32F103C8T6 on a Bluepill board with STM32F303CCT6 for initial porting, before receiving the NanoVNA PCB from Hugen. For debugging, I use ChibiStudio, OpenOCD,
and a Nucleo board as ST-Link v2. Becasue DFU doesn't work at F303 yet, I have to use ST-Link to burn firmware, instead of USB DFU. Maybe a bootloader is needed for F303, like m0nka's mchf SDR transceiver.

For larger LCD, I tried a 3.5" 480x320 LCD with ili9488 controller on my F303 Bluepill. I was able to port to ili9488. But since ili9488 doesn't support RGB656 as ili9341 does, each pixel needs 3 bytes to trasnfer RGB666 or RGB888. And since STM32 doesn't support SPI with 24-bit data, each pixel takes three times of one byte transfer; while ili9341 just neesd to transmit 16-bit data at one time. Besides, the 3.5" LCD has twice the pixels to draw than the original 2.8" 320x240 LCD. Even after some optimization, the result is still dismal. It seems that ili9486, which supports 16-bit per pixel over SPI, would be a better solution than ili9488.

With larger LCD and SRAM, I was hoping to increase the number of sweep points from 101 to 201. Athough I was able to double the sweep points, I could not save the calibration result. Each time I save the calibrated data, it hangs. I think I need to look into the flash.c code.

Ken

Hello Kurt,
that is a wonderful description of what you three did a while go when measuring the performance of ferrite materials. Very cool stuff and because I'm just looking in exactly that topic I have a few questions questions to your fixture. I'm still on a "beginner level" using VNA for measuring ferrite impedances, so sorry for any stupid questiosn.

When you have inserted the ferrite bead and turning the adjustment knob it will contact the SMA F connector at some point. What type of SMA connector did you use? I would imagin that the inner conductor gets damaged pretty fast after using the fixture a few times? Or is that a "special" type of SMA with more robust inner-conductor design ?

Do you know if something similar is commercially available?

Secondly, I like to characterize the ferrits with DC-BIAS current from 0 up to 10 Amps.. That changes ferrite impedance and frequency, depending on saturation of the material.
See
For relatively large ferrite cores it might be easy just adding an additional wire through the inner dimeter, with a DC-current. That way the RF- and DC wires are separated.
But for smaller components like wound-beads (e.g. Fair-Rite 2961666671) I need to "inject" the DC current into the signal line of the VNA. Do you have any experience how to do that best? What kind of RF-filtering / DC-blocking needs to be build ? How is calibration performed then? The additional filtering will definitely change the overall S11 response, but I want to see only the ferrite effect on the Impedance and not any effect of that filter. The intended frequency range is up to 500 MHz for #61 material, but ideally for full span of nanovna up to 1 GHz.

The coordinate system of the Smith Chart is a plot of complex numbers representing
the reflection coefficient. The reflection coefficient is basically a ratio, and thus has no units.

You can read reactance and resistance from the arcs and circles that are drawn as part of the Smith Chart.
The Smith Chart itself is plotted in the complex plane, the coordinate grid is not typically shown on most
Smith Charts but could be either standard issue cartesian coordinates or polar coordinates as fully
described in Jeff's tutorial. That is the whole point of his tutorial. He makes it very clear what and how
stuff is plotted on the Smith Chart. Most Smith Chart tutorials just give the mechanics of how to use it
with a very brief mention that it involves the reflection coefficient.


Here is a brief section of the part where he describes what is being plotted:

Its impedance mismatch (as represented with its Reflection Coefficient) from our target impedance of Zo = 50 ohms calculates to be:
Γ = 0.8 + j0.4
And we can plot its position (using Cartesian coordinates) on the complex plane:

The only thing I might add to the tutorial in this regard would be to state that the reflection coefficient has no units.
Other than that, any suggestions out there on how to improve this aspect of the tutorial?
It is a key concept, and quite confusing.

Jerry, KE7ER


On Sun, Oct 20, 2019 at 01:46 AM, Oristo wrote:

a quick skim read indicated no reference to what the x & y axes are,
which is something I have noticed with several descriptions of the Smith
Chart.

Sad, because Smith chart simply plots S11 reactance vs resistance.

On Sun, Oct 20, 2019 at 12:01 AM, Dr. David Kirkby from Kirkby Microwave Ltd wrote:

On Sun, 20 Oct 2019 at 01:08, Jerry Gaffke via Groups.Io <jgaffke=
[email protected]> wrote:

Don’t you feel that the Smith Chart is *far* away from the original posters
question about things for *beginners*? The original poster suggested
finding the resonate frequency of an antenna.

Hi Dave. I was replying as an aside to Dick’s post-script aside regarding Smith Charts in the thread, here: /g/nanovna-users/message/5216

Best,

- Jeff

If you go through each cal step and save, why can't it simply over-write what
was in the memory in the first place?

My >>guess<< is limited code space and performance:
* Typical use wants correction for each reading
- don't want custom loop for calibration
- don't want special test for calibration to slow that reading loop
- before calibration, set corrections for identity

Hi vk2krn -

"RESET" on touch screen only clears current calibration data, does not affect SAVE 0-4
- unless by some firmware bug[s]
~ configuration data (including calibrations) is liable to accidental damage by firmware bugs.
~ this has caused nanoVNA hangs at power on and other times, loading damaged settings.
~ such hangs doom firmware DFU menu option

Now if I make a fresh calibration for a different frequency range,
(different from C0 which is my default)

That is my default; tweaked nanoVNA has nearly 15 octaves (2**(15) == 8**5),
so 50-400kHz, 400-3200kHz, 3.2-25.6MHz, 25.6-204.8MHz, 204.8-1500MHz

is it correct to just carry out the full cal without selecting "Reset"

No, you should RESET before each cal
(IMO, firmware should change to automatically reset in first cal step)

and saving the new Cal data to a new memory location?

Consider that each of 5 possible stored cal datasets is
101 frequency samples for inline and quadrature phases of reflected and thru.
There is a >>current<< correction dataset (probably in RAM)
that may or may not duplicate one of RESTORE 0-4, depending on current situation.
Firmware has limited storage available; I >>suppose<< that:
- it lacks space and/or code to store fresh samples without applying whatever is current correction
- cal procedure processes fresh samples as if uncorrected
- without RESET, >>double<< processing occurs, with invalid results.

I personally do not understand the need for "reset". If you go through each cal step and save, why can't it simply over-write what was in the memory in the first place? If you save the data with "save" then why do you need to reset anything. What is "reset" actually doing to enable the saving of new cal data?

Perhaps the "reset" is doing something that could better be explained with a different word?

NooElec is more like 1 person not 50! been around for a while. Had a very bad warranty experience with them.
I do not see any advantage in ordering from them.
Authorized give me a break!
Tim

--
Tim W4YN

Hi Starsekr and All,

Another test jig for measuring Ferrite is shown here:

There is also a section describing process for plotting Complex Permeability using SimSmith. The example compares measured complex permeability to Fair-Rite published data.

John KN5L

Before Smith charts, my gnuplots were smaller by gif than png,
and I did not think to recheck.

OK, gnuplot Smith chart is slightly smaller by gif than png on my Windows PC,
but gif takes a disturbingly long time to generate, so updated:



.. with a small 'png.p' gnuplot script
for generating "gnuplot.png" from current plot
(remember to rename it so not overwritten by next 'png.p' invocation)
and added Roman's improvements to 'smithlabel.p'

#60 : @ Gary O'Neil - 17 October 2019:
/g/nanovna-users/message/5075

Dear Gary,

Allow us, please, to reply your most valuable comments and suggestions
as follows:

C1-GN : 1st Paragraph : "Your #51 full" ... "you have presented."

A-G&Z : On the known and most understandable difficulty of following links :

- Are you sure that you really read instead of #50:

/g/nanovna-users/message/5061

its online content, contiguously including all the Figures, here:


61 - [NanoVNA] ~ [ VNA] - Πω? Κ?ναμε την Σ?γκριση - 17 Οκτ 2019, 00:57

where, except just three 3 non-English, title and salutation lines which you can
ignore of course, there are only a few, non-essential, links ?

Why do you not save this html page for comfortable off-line view ?

C2-GN : 2nd-4th Paragraphs : "I believe you presented [... ] of being FACUPOV"

A-G&Z (a) : On the [openwatcom-fortran][1.9]:

- What do you mean by "shadowed" ? The link at:

ftp://ftp.openwatcom.org/install/open-watcom-f77-win32-1.9.exe

is alive and this download of 66,473,025 bytes is still a working one.

A-G&Z (b) : On the FORTRAN in general:

- But, if you do mean that FORTRAN is no more "in trend" or "in fashion",
then there is only a little that we can do about it and definitely to not
dispute your points : 1) - 6) , which are simply a matter of taste.

C3-GN : 5th-10th Paragraphs : "If I may suggest [...] Linux version"

- Thank you very much ! We can assure you that we already took into
account all that, as it may be reflected, as we hope, in the messages
we sent meanwhile to the group.

Best regards,

gin&pez@arg

60#

@Oristo - Is my understanding correct that if I select "Reset" and then proceed with a calibration, ALL previously stored Cal data is cleared? Now if I make a fresh calibration for a different frequency range, (different from C0 which is my default) is it correct to just carry out the full cal without selecting "Reset" and saving the new Cal data to a new memory location?

Here is an example for gnuplot line colors:

Hi Rudi

First of all the use of the testcables from the NanoVNA female SMA adaptor (the SMA male female adaptor saver) to the reference plane is a bad idea. The female female adaptor used during calibration has loss and has influence on the sweep of the semirigid cable sweep so the sweep run outside the Smith chart. Adding the female female adaptor at the end of the semirigid cable is also wrong as the loss in the female female adaptor again creates a wrong trace around the Smithchart. The female female adaptor supplied with the NanoVNA has in particular high loss and any other female female adaptor has some loss to account for.
So do the following:

Directly at the VNA SMA female Ch0 adaptor do a S11 calibration (do not use the supplied open adaptor) and connect you semirigid cable directly afterwards and leave the far end unterminated. Ideally the semirigid cable should end blindly but the SMA male adaptor has little effect loss wise, but it is there despite small and hardly detectable. The big bug is the female female adaptor you must not connect as you did.
When dealing with the N calibration kit then use a SMA male to N female adaptor and at first directly connected to the NanoVNA Ch0 female adaptor. Afterward you may experiment with adding test cables and do the calibration from there where you are forced to use the female female adaptor or better a SMA female to N female adaptor.

Use the NanoVNA saver feature to save the first sweep directly on Ch0 to "Save as Reference" as you then can compare step by step what you are doing.

Do not expect the NanoVNA to beat a more advanced VNA, it still has some shortcoming and you got a version which is characterized by some as a bad clone ?

Kind regards

Kurt



-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af reuterr@...
Sendt: 20. oktober 2019 10:02
Til: [email protected]
Emne: Re: [nanovna-users] The Male Calibration Kit for AliExpress



Hello Kurt,

Thank you very much for taking care about my question.

I will write my answers below your questions.



On Sat, Oct 19, 2019 at 10:56 PM, Kurt Poulsen wrote:

1. Did you use a male SMA to N female adaptor directly fitted to the

SMA

No.

female adaptor for Ch0 ?

Just a male-female thread saver, as could be seen on the photo.

2. Did you before using the NanoVNAuser ensure the NanoVNA was

correctly calibrated directely at the NanoVNA Ch0 female adaptor with

the supplied calibration kit ?

The nanoVNA was calibrated at the *Reference Plane* shown in the photo.

The used *Calibration Kit* was the supplied one.

The *Open* SMA part was replaced with a *female-female adapter*.

2a. and did a sweep of a 25 to 50cm SMA test cable show better result?

I can not understand the question.

3. The test cable is probably too long and shall be a N type cable

with no adaptors fitted

I do not have that.

4. Your NanoVNA has a poor bridge as NanoVNA as far as I know has no

internal error correction

I measured the 50 Ohm resistance with a 3 VDC supply and a 4 digit *Ampere meter* and *Volt meter* The nanoVNA CH0 port shows 49.23 Ohm.

The 50 Ohm load shows 50.72 Ohm.

A self made 50 Ohm load from 2 SMD 0805 100 Ohm 1% resistors in parallel shows 49.99 Ohm (interesting).

There is probably more reasons and it would be better if you explained

in details you setup

Please see the attached photo: nanoVNA_50cm-RG402-open2_f-f_DSC08152.jpg

I have attached 2 screen captures from nanoVNA-saver with the S11 measured 50 cm RG402 coax cable.

nanVNA-Saver_50cm-RG402-open2.png - End of RG402 SMA male plug open nanoVNA-saver_50cm-RG402-open2_f-f.png - End of RG402 male plug with *female-female adapter* It is interesting how much the *Parallel X* capacity changed.



The nanoVNA-saver has an additional calibration with 5 segments.



What can I enter in the window Calibration->Calibration standards in order to improve the measurement?

73, Rudi DL5FA

Hi Rune -

I just tried it with a different plot I have a script
making hourly plots, 1024x768 pixels. .gif is 18547 bytes, .png is 14244
bytes. So I definitely see .png being smaller.

Oops, thanks for testing. Before Smith charts,
my gnuplots were smaller by gif than png,
and I did not think to recheck.

On Sun, 20 Oct 2019 at 11:12, Oristo <ormpoa@...> wrote:

set terminal png size 1000,1000

gif will be smaller than png

That's interesting. I just tried it with a different plot I have a script
making hourly plots, 1024x768 pixels. .gif is 18547 bytes, .png is 14244
bytes. So I definitely see .png being smaller.

(Also, if you set it to gif, you end up with a gif file, which you probably
wouldn't want anyway. ;-)

--
Rune / 5Q5R

i cannot save correct cal data to any location other than C0

Do new calibrations not work immediately after creating?

Run RESET before any calibration or to run uncalibrated.

nanoVNA always loads C0 (if saved) at power-on;
use RECALL to load any other saved calibrations.

WSL doesn't have X Windows support

VcXsrv works for me with WSL since 2017

set terminal png size 1000,1000

gif will be smaller than png

set linetype 1 lc rgb "gray60" lw 1 pt 1
set linetype 2 lc rgb "red" lw 2 pt 1

Thanks, I forgot to update color coding.

gnuplot works fine directly from Windows cmd.exe


I use a bat file to set a custom PATH when using gnuplot from cmd.exe:

C:\Users\ormpoa>type gnuplotPath.bat
path=C:\Program Files (x86)\gnuplot\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\WindowsPowerShell\v1.0

Hi Rune,
thanks for your kind offer, S21&S11 on the same plot is for sure good enough!?
Kind regards
Norbert, DG1KPN

Don’t you feel that the Smith Chart is *far* away from the original posters
question about things for *beginners*? The original poster suggested
finding the resonate frequency of an antenna.

Smith charts show resonances, where plots crosses horizontal axis.

a quick skim read indicated no reference to what the x & y axes are,
which is something I have noticed with several descriptions of the Smith Chart.

Sad, because Smith chart simply plots S11 reactance vs resistance.