开云体育

On Fri, 27 Sep 2019 at 20:00, yza <yzaVNA@...> wrote:

19 : "true value" - also : @Dr. David Kirkby :

/g/nanovna-users/message/3207

Hello,

We both thank you very much for the time you spent
to compose such a lengthy reply, by which you definitely
assured us that you are really interested on this subject,

I am interested. As an undergraduate student I spent some of my time at a
calibration laboratory run by the Ministry of Defence in the UK. That
placement got me interested in metrology, so measurement uncertainly.
However, I am *not* a metrologist. I have an interest in the subject, but
are not an expert.

Core of Measurement Uncertainty in VNA/nanoVNA,
that is, for example, when he insists to force his mind
to be trapped by the conventional triplet of the
"standards" (S,L,O), instead of the most general
one of "loads" (A,B,C)
-
he has been warned.

I can assure you that I am *not* trapped into assuming that the calibration
standards need to be open, short & load. I have designed and sold waveguide
calibration kits, where it is totally impossible to make a “open “. In
fact, I wrote an explanation of why an open can not be used for waveguide
calibrations about a year ago. You almost certainly know why, but I will
post the link, hoping to convince you I am not stuck into this
open/short/load procedure, and perhaps it will be useful to others



However, I am trying to bring to your attention the fact that in the
English speaking world, your use of “loads (A, B, C)” is *very confusing.*
Many people will interpret that as meaning three resistive devices.
Instead I believe that you should use the term

“*calibration standards (A,B,C)”*

Personally, I would write something like

“calibration standards (A, B, C), where typically A, B and C are a short,
open and a load, but A, B & C can theoretically be any combination of
three devices which have a different reflection coefficient at every
frequency over which they are used“

Although I admit that it is longer to write, it has several advantages.

* It avoids the use of the word “loads” which is confusing in the English
language.
* Making a reference to the three calibration standards typically being a
short, open and load will help the reader understand more.
* Stating that theoretically the calibration standards can be any
combination of three devices having different reflection coefficients, will
educate some readers who may not know this.
* By stating that theoretically any combination of three devices can be
used, will allow you to address the problem I stated earlier, that if the
phase of the calibration standards are too similar, the calibration will be
poor.

I am really trying to help you, but unless you address some of my concerns,
in particular the need to write your work in one document that at least
some English speaking readers will be able to follow, I fear nobody on this
group will understand your work.

I have a Fortran compiler (gcc), and probably could compile your Fortran
code. However, I am not going to do this unless you can give me a clearer
explanation of the scope of your work.

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United
Kingdom

The NanoVNA firmware (as I understand it) assumes that the SMA Short, Open, and Load standards are ideal, with the single exception of defining the Open to have a C0 50 femtofarads.

I was wondering -- is this a sufficient definition for these standards, given the NanoVNA's frequency limit of 900 MHz? Or should we also be including values for C1-C3, L0-L3, Offset Loss and Offset Zo?

So I wrote some Matlab code to calculate a standard's Reflection Coefficient (Gamma) based on Keysight's "Full" model (using C0-C3, L0-L3, Offset Loss, Offset Delay and Offset Zo), and compared this to a simplified model in which all of these values were set to 0 except for C0 and Offset Delay.

I've attached a table of results.

My conclusion is that the simplified model (C1-C3, L0-L3, and Loss are set to 0, and Zo is set to 50) is perfectly adequate for use with the NanoVNA.

But my math could be wrong. Do my results seem correct?

(A detailed explanation of what I've done is in the top post of this blog: )

Thanks,

- Jeff, k6jca

P.S. I did NOT zero-out Offset Delay in my simplified model (I believe (but have not verified) that this delay IS assumed to be zero in the NanoVNA firmware). If the actual Offset Delays for the Open and Short standards are defined by their supplier to be non-zero, then the net effect of zeroing out these delays is to simply move the reference plane of each -- i.e. rotate the Open and Short's reflection coefficients around the Smith Chart circle. However, note that zeroing out the two delays could become an issue if the Short's delay is NOT equal the Open's delay -- you would then lose this difference in delays (should this delta be large enough to be important)).

I conducted a small test with IP5303.
IP5303 does not work longer than 30 seconds with a load that consumes less than 65 milliamps. Nanovna consumes 90-110mA.
With such a current consumption, IP5303 should power the device for a long time, until the battery energy is depleted.
Test with the new IP5303, connect a 47-38 ohm resistor parallel to the capacitor C47.
If IP5303 will turn off at a load current of 100mA, then IP5303 is fake, or the elements in its bundle are faulty, although there are few of them.

Hi,

That's for you :

@edy555 edy555 released this 2 hours ago
in-device TDR support (contributed by @cho45)
add scan command for multisegment scan excess 101 points
fixed invalid sweep at the narrow span (<5kHz)
fixed failures caused by a race condition between USB and measuring loop
find device automatically in python script




Regards,
David F4HTQ

digit255,

Larry addressed this topic a couple of days ago:

" Larry Rothman
Sep 26 #3138

As hugen mentioned yesterday, he is thinking of re-integrating (merging?) the larger fonts with his current source code.
The larger fonts were created in order to display Chinese characters.
Hopefully, that will offer more options to all us older, more mature folk."

Basically the larger font size cannot be used with the most recent firmware until memory is freed up by first dumpling the space used to create Chinese characters.


Herb

Hi Rune,
I tested a BP-filter and I am wondering why the VSWR plot is interrupted. If you take a look at the attached screenshot, there is no VSWR line from about 268MHz till 846MHz. A tested LP-filter shows a similar symptom.
Would you please take a look at it?
Did anybody of the group see similar VSWR plots?

Kind regards
Norbert, DG1KPN

Not yet but forum member hugen is looking into something to help with this.
Stay tuned...may take a week or two.

Bo

I am the owner of RFzero (THANK YOU!) and of three, soon to be four, Nanovnas. While it does not fundamentally alter the discussion here, The Nanovna only has two frequency ranges: 55kHz to 300 MHz is accomplished using the fundamental. 300.01 MHz to 900 MHz is accomplished using the third harmonic of the Si5351 with the fundamental level boosted about 9 dB for this range.

Nanovna uses a second 5351 oscillator as a "tracking" source that operates 5 kHz above the measurement frequency and is used as a heterodyne source to the mixer which generates the 5 kHz audio signal for measurement. I believe (I have measured it but cannot find the data) that Nanovna uses the third harmonic for this "tracking" source from 300 - 600 MHz and then switches to the 5th harmonic for 600-900 MHz.

So it goes like this:

55kHz - 300 MHz: Measurement = Fundamental; "Tracking" = Fundamental

300MHz - 600 MHz: Measurement = 3rd Harmonic; "Tracking" = 3rd Harmonic

600MHz - 900 MHz: Measurement = 3rd Harmonic; "Tracking" = 5th Harmonic

BTW..... I have a HUGE problem trying to decide between RFzero and Nanovna as which is my favorite instrument!

WA8TOD

Rune wrote:

" I'd have wished they'd contact me about it instead, but as it's open
source, they are of course free to use the code.

I may need to make sure they don't reuse the name, though, so there isn't
confusion about which version is which "
---------------------------------------------------------------------------------------------------------------------------------

I remember edy555 having similar thoughts after Hugen started selling the NanoVNA based on his original open source design. For a brief history lesson, see posted comments from edy555 to Hugen below:


" Hi hugen,

I'm an original developer of NanoVNA.

When I knew that you made the clone of my project, I very surprised that
you made PCB from schematics, and as an enthusiast, I felt pleased with
your challenges such as frequency range expansion and your original pc
software.

However, I was annoyed that you sell your clone without any prior notice
to me. Though I had a plan to forge and sell my product, it becomes
difficult.

Furthermore, your act cause that your design material was stolen by other
clone makers, and quite many units were sold in aliex, ebay and also amazon.
Those include worse clone as you say. This is a worrying state of affairs I think,
and you might agree.

To distinguish any unexpected clones, I propose that you should change your
product name by adding a suffix such as NanoVNA-H from your name, and it
should be shown in the market and printed on your product.

And also I'd like to publish the name of qualified products, so customers
become able to avoid worse clone.

I hope you think about this issue seriously.

Regards,

edy555
Aug 13, 2019 "


Herb

Is there build firmware with battery indicator, 1500 MHz and big font ?
I only found with small font.

Hi Herman,
That is really strange. However, your Nano is not lost -just buy a cheap power bank that puts out 5v and use that to power the Nano, which really runs from 5v, not 3.7v.
In fact, you can always buy a very inexpensive and small charger-inverter board from the likes of banggood for around $2.
Yes, I know it's a pain in the butt but at least you can keep using your unit.

These boards use the 54xx chip..




Note both boards use a 4R7 coil but the cheaper board uses a 33ohm in series with USB 5v but the more expensive board uses 2 - 1ohm in parallel for current limiting.
I wonder if these chips are sensitive to input current from USB and are being damaged somehow.
Try placing a 33ohm at the 5v input to the chip and see if that helps.

You might be able to transplant parts from one of these modules to the Nano.

Regards
Larry

Its time to give Rune a cup of coffee!
Where is the paypal button?? :-)

Peter
Op 28-9-2019 om 15:06 schreef Rune Broberg:

On Fri, Sep 27, 2019 at 05:38 PM, yza wrote:

where, the sine qua non Core Uncertainty is due both to:

(a) the 5 = 2 + 1 + 2 uncertainties of the S, L, and O standards,
which are known from their manufacturers' data, as well as

(b) the 8 = 2 x 4 inaccuracies of the four VNA readings,

that is the Core Uncertainty is due to a combination of
26 = 13 x 2 error bounds, in total.

Hello yin&pez,

Thank you for introducing the topic of errors of error-models. This discussion is very interesting.

To help me understand, I would like to ask a few questions.

First, you mention that the S and O standards have 2 uncertainties each (while the load has 1 uncertainty). How are you defining these uncertainties? Are they defined as uncertainties in a Standard's Reflection Coefficient (that is, uncertainties in rho and theta), or are they defined to be something else? And why does the load only have 1 uncertainty?

Second, you mention the inaccuracies of the four VNA readings (two inaccuracies for each of the four readings, which I assume are the three readings for the SOL standards plus a fourth reading for the device-under-test). Could you please explain what these inaccuracies are and how one determines them?

Finally, how did you arrive at a Core Uncertainty of 26 error bounds? I can see how you calculate 13 (by adding 5 plus 8), but I do not understand why this value is then multiplied by 2.

Thank you very much for your time.

Best regards,

- Jeff, k6jca

22 : We Learn Our Lessons - NanoVNA Calibration
Considerations and Procedure_v1.1

@Gary O'Neil : /g/nanovna-users/message/3259

Hello,

Thank you very much for your time that you spent to
subjectively but kindly advice us lengthy about the
existed perfect climate that rein among the honorable
members of this forum !

We much appreciate that; indeed. Therefore, allow us,
please, to inform you that, as we already assured at:

18 : @Dr. David Kirkby : /g/nanovna-users/message/3192

and we can also gladly repeat here:

"
we always try to be good listeners, we shall also attempt
to follow your valuable subjective suggestions on this very
subject and to learn our lessons : Thank you once more !
"
On the occasion, allow us, please, to ask you to accept
our respects for your Excellent Work you have done into
the proven so valuable to us:

"NanoVNA Calibration Considerations and Procedure_v1.1"

where we read that :

"This document is mostly the work of Alan Victor, W4AMV
and his colleague Gary O’Neil, N3GO":
/g/nanovna-users/files/NanoVNA%20Calibration%20Considerations%20and%20Procedure_v1.1.pdf

Thanking you again,

Sincerely,

gin&pez@arg

I'd have wished they'd contact me about it instead, but as it's open
source, they are of course free to use the code.

I may need to make sure they don't reuse the name, though, so there isn't
confusion about which version is which.

Thanks for letting me know :-)
--
Rune / 5Q5R

Rune,
The popularity of NanoVNA Saver continues to grow. I just saw this over on the NanoVNA-F user group.

"We are currently modifying nanovna-saver (a good NanoVNA PC software) to fit NanoVNA-F, so stay tuned"

Good luck with them trying to keep up with your pace of development in a different branch :)

Herb

21 : The Mathematica Code and the Unpacked DERDEI Code

@Jose Luu : /g/nanovna-users/message/3278

Hello,

Thank you very much for your kind interest in our work !
We much appreciate that; indeed.

Especially, the fact that you spend your time to read our
documentation for our code. Therefore, allow us, please,
to inform you that, unfortunately enough, we are not in
place to fulfill your inquiry about our mathematica code,
which is already a more than ten 10 years old code ,
for an hierarchy of numerous reasons:

From the most generic first top one, which has to do with
the fact that meanwhile we radically changed our point of
view regarding proprietary software and thus we don't want
to support any one, in any way, even in our humble one,
anymore, down to the most specific last bottom one, which
has to do with the fact that this code is nothing more than
yet another example of the worst of programming style,
of a most personal character, that confused even us some
years ago, when we tried to find out what we had wrote
those days...

Well, after all that subjectively said, we conclude with our
definite decision : this mathematica code of ours is
definitely an unworthy, definitely a non-presentable,
definitely a not for publication code.

However, allow us, please, to also inform you that your inquiry
motivated us to decide that it is really now the time to unpack
the DERDEI code and distribute it, also under an MIT /F/L/O/S/S/
license, thus, we would like to please you to give us some time
to prepare the unpacked code in a form that will allow us to
decide that it is perhaps a somehow worthy, somehow
presentable, and thus a definitely publishable code.

Sincerely,

gin&pez@arg

21

Thanks for the info Warren,
73
de VU2PGB
VEEN

An update on dynamic range. In the photo the upper unit is the unshielded "worse" version and the bottom is the latest, best, shielded version. Both are calibrated for the 600 - 900 MHz range, Logmag, Ch1 with Through calibration. The reference level is the seventh level on both.

The diffenence in noise floor ranges from 3 - 10 dB across the range. Both units meet the dynamic range spec of 40 dB with the shielded unit showing better than 50 dB at the low end.

For my purposes the two units are substantially identical.

WA8TOD

is this modified software available to other users?

Gyula