开云体育

Very helpful, Dave, thanks much.
73, N0AN

I have created some specific cal files for:
100-500 MHz (evaluate 2 port duplexers 2m/70cm)
50 kHz - 500 MHz for Generic Use
400-500 MHz to look at filter insertion loss

All but the last one is doing what I want. I'm not sure the nanoVNA can
measure insertion loss of < 1 dB , no matter what settings I use. Each cal
is based on 10 bins, so I have 1000 points.

73, N0AN
Hasan


On Mon, Sep 14, 2020 at 11:59 AM David Eckhardt <davearea51a@...>
wrote:

You are better off using a larger number of 101 sample 'bins'. With a cal.
from 50 kHz to 1 GHz and using only 1 "bin", you have only 101 points
across that sweep. That would amount to a measurement of the cal and data
every 9.9 MHz. The resolution with that cal applied to any 2-meter or 70
cm filter will be unacceptable. If you specify 10 bins of 101 points each,
that will yield a measurement and cal point every 0.99 MHz. Still pretty
bad resolution. It is not a trick of the display, but the number of actual
cal'ed and measured points. So, to get a meaningful display and
measurement, its better to define a much smaller sweep appropriate to what
you are after and assess the granularity of the data to set the number of
'bins' over that sweep.

The result of your comparison between the duplexer and triplexer may well
be correct. With the triplexer, you have the added roll-off contributed by
the center BPF to add to the high-frequency skirt of the 2-meter LPF and
the low-frequency skirt of the 70-cm HPF. That likely accounts for the 8
dB difference between the two designs.

Dave - W?LEV

On Mon, Sep 14, 2020 at 9:31 AM Hasan Schiers N0AN <
hbasri.schiers6@...> wrote:

I have a nanoVNA-F that I am very pleased with.
I have done the internal calibration.

I use nanosaver software on Win10, and am also very pleased with it. Do I
recalibrate using the nanosaver software for every different band filter

I

am testing and save that calibration using the nanosaver software?

1. To measure the insertion loss of a bandpass or highpass filter that is
designed for 70cm, do I recalibrate using nanosaver to the design range

of

the filter, or use the very wide range that was set to my internal
calibration of the unit.?

If I want to know how much loss the 70 cm. filter has inside it's

passband

or targeted use frequency, should I recalibrate using 400 to 500 MHz

sweep

range, instead of the 'base' calibration I did which used 50 KHz to 1000
MHz?

On the other hand, when I want to evaluate a 2m/70cm duplexer's
performance, I'm not so much looking at insertion loss as I am

attenuation

of the 2m signal by the 70cm port (to prevent desense of the 2m receiver

by

the 70cm uplink transmitter). I am seeing about 56 dB of rejection by the
duplexer's 70cm port, at 145.9 MHz (Non-used 2m port terminated in 50
Ohms.

I am also seeing about 8 dB more rejection of the 70cm uplink on 2m

using a

Triplexer (HF/2M/70cm) LP-BP-HP, as a filter instead of the duplexer
(LP/HP). Is that expected behavior?

2. The default value in the Sweep control for what appears to be

resolution

is 1, I have played with various values from 1 to 10. Obviously, the
display is much smoother with larger numbers, but it this a change in
resolution (of the data measurement) or is just a display trick? Do I get
better precision using a number greater than 1?

Thanks for your help, this little unit is a ham's dream come true.

73, N0AN
Hasan

--
*Dave - W?LEV*
*Just Let Darwin Work*

On 9/15/20 2:32 PM, Colin McDonald wrote:

Excellent! Thanks for the clerification. I didn't really understand the difference between MA and RI before.
I read through the touchstone specifications document, but don't remember if there is a set standard for s1p/s2p with regard to which contains RI and which contains Ma data. It seems to me that it would be nice to be able to determine which you wish to export from within the program then export as such.

you look at the header of the file starting with #

! Vector Network Analyzer VNA R2
! Tucson Amateur Packet Radio
! Saturday, 9 November, 2019 17:48:47
! Frequency S11 S21 S12 S22
! ListType=Lin
# HZ S RI R 50

^^

tells you it's in real/imaginary form.


here's a MA format one

# MHZ S MA R 50
10.0000 0.915 -63.193 57.033 142.886 0.013 51.862 0.525 -89.730
20.0000 0.830 -99.758 40.867 122.518 0.021 32.533 0.638 -123.891
30.0000 0.787 -121.018 30.176 110.861 0.022 20.814 0.684 -140.388
40.0000 0.765 -134.056 23.331 103.406 0.023 15.181 0.708 -149.561


And, there's a third format.. "DB", where the S parameter is in dB magnitude and degrees phase.


# MHZ S DB R 50
!Frequency S11 dB S11 DEG S12 dB S12 DEG S13 dB S13 DEG S14 dB S14 DEG
! S21 dB S21 DEG S22 dB S22 DEG S23 dB S23 DEG S24 dB S24 DEG
! S31 dB S31 DEG S32 dB S32 DEG S33 dB S33 DEG S34 dB S34 DEG
! S41 dB S41 DEG S42 dB S42 DEG S43 dB S43 DEG
S44 dB S44 DEG
.3000 -2.335942E+001 9.800809E+000 -6.468416E-001 -3.088451E-001 -1.268429E+001 8.693548E+000 -1.267583E+001 -1.722793E+002
-6.292930E-001 -4.186788E-001 -2.338740E+001 7.548462E+000 -1.271809E+001 -1.711522E+002 -1.271509E+001 7.882564E+000

Unless there is a standard already in which case you just save to s1p or s2p depending on which data you want.

S1P is "one port" data, S2P is two port. S4P is four port (you see this for transformers with a center tap)..

Thanks for helping so far. I will have more questions.
Regards
Colin va6bkx
-----Original Message----- From: Jim Lux
Sent: Tuesday, September 15, 2020 7:12 AM
To: [email protected]
Subject: Re: [nanovna-users] Writing nanovna results to csv file #nanovna-saver
On 9/15/20 3:55 AM, Piero Tognolatti wrote:

Hi Colin,

Please, note that s2p file is in the "RI" format, your formula to compute SWR is not correct.?? "RI" means that one column (2nd column) shows the real part of S11, while the other column (3rd column) the imaginary part of S11.

Your formula is correct when the s2p file is in the "MA" format, which means that one column is the Magnitude of the S11, while the other column is the Angle of the S11. Let me recast your formula saying that SWR = (1+M) / (1-M)?? where M is the magnitude of S11, taken from the proper column of a s2p file in MA format.

At the moment I cannot run nanoVNA-saver, so I don't know if the exported s2p file are in MA or RI format.

regards

Piero, I0KPT

if it *is* in real/imaginary form, then
SWR = (SQRT(R^2+I^2)+1)/(SQRT(R^2+I^2-1)

Il 14/09/2020 22:51, Colin McDonald ha scritto:

With some more research I managed to create a formula to show swr in a SNP spread sheet.
SWR = (MA+1)/(1-MA)
Where MA (or RI) is the second colomn in the spread sheet.
for instance SWR = (B3+1)/(1-B3)

Excellent! Thanks for the clerification. I didn't really understand the difference between MA and RI before.
I read through the touchstone specifications document, but don't remember if there is a set standard for s1p/s2p with regard to which contains RI and which contains Ma data. It seems to me that it would be nice to be able to determine which you wish to export from within the program then export as such.
Unless there is a standard already in which case you just save to s1p or s2p depending on which data you want.

Thanks for helping so far. I will have more questions.

Regards
Colin va6bkx

On 9/15/20 12:35 PM, WB2UAQ wrote:

To Helmut Wabnig, I experimented with the TDR measurement using Saver and found that the lengths appear to be where the first impedance change takes place. The Nano doesn't send down the transmission line a very narrow pulse like a true TDR. A true TDR will show not just the first abrupt Zo change but subsequent bumps. I checked this stuff out just a few weeks ago. With 15 ft of RG-8 connected to 90 ft of 75 ohm cable attached to an antenna. Using a pulse generator and a scope to create a TDR, I could see each bump....where the RG-8 reaches the 75 ohm cable and then where the 75 ohm cable reaches the antenna. The Nano reports just the first transition from RG-8 to the 75 ohm stuff. I must say that it was pretty accurate. As long as the freq accy is reasonable the first bump should be pretty good using the Nano.
Please....others add their experience I only used the Nano for its "TDR" function for a short time.

and, you will see all the other bumps, if you can get a "time domain" display. I suspect the "cable length" function looks for either the first "big" discontinuity, or the one with longest delay.

To your query, for research use only, attached papers from 1924 (background), 1932 and 2010 to show where things evolved.

You can also find info on the radiation resistance issue in Kraus, Jordan and Balanis.

Best regards,

AG6CX

To Helmut Wabnig, I experimented with the TDR measurement using Saver and found that the lengths appear to be where the first impedance change takes place. The Nano doesn't send down the transmission line a very narrow pulse like a true TDR. A true TDR will show not just the first abrupt Zo change but subsequent bumps. I checked this stuff out just a few weeks ago. With 15 ft of RG-8 connected to 90 ft of 75 ohm cable attached to an antenna. Using a pulse generator and a scope to create a TDR, I could see each bump....where the RG-8 reaches the 75 ohm cable and then where the 75 ohm cable reaches the antenna. The Nano reports just the first transition from RG-8 to the 75 ohm stuff. I must say that it was pretty accurate. As long as the freq accy is reasonable the first bump should be pretty good using the Nano.
Please....others add their experience I only used the Nano for its "TDR" function for a short time.

Yeah, had to return my eBay NanoVna-F with its malfunctioning menu for refund. After discovering others are experiencing similar problems thus putting reliability/life expectancy in question, I've decided I can live without it.

to whom it may concern - we still make
the required manipulation and graphics
of the valuable NanoVNA multi output
raw data (with its initial, already one year
old firmware version, of course) using
reliable external programs - only rarely,
if ever, we need to compare our results,
we get in this way of use, with those
of its direct output - sincerely

I've found that occasionally, my NanoVNA (or, more properly, my NanoVNA + NanoVNA-Saver) doesn't come back with a reasonable length on the coax.

Invariably, it has been that I inadvertently set a frequency measurement range that was wrong, or for which there wasn't valid calibration data.

The TDR functionality is done by doing a Fourier Transform of the S11 data, which turns it from frequency domain to time domain.

The *resolution* in time domain is the inverse of the total bandwidth of the input data (so sweeping from 10-20 MHz gives you 10 MHz bandwidth, which is 100 ns resolution)

The max range in time domain is set by the spacing between frequencies. If you have 1 MHz steps, sweeping from 0-100 MHz, you have 1 microsecond "max range" and 10 ns resolution. It's two way, so 10ns of time is actually 5 ns of one way time, which is about a meter of cable with typical velocity factors.

I'm not sure how NanoVNA-Saver deals with the calibration when doing the enhanced "multiple sweeps to cover a wider band".

But if you calibrated for, say, 0-50 MHz, and then switched to 0-100 MHz to sweep for the TDR, I think it just extrapolates beyond the 50 MHz to cover the uncalibrated area - and that can be *very strange*, depending on how the extrapolation works.

You can also get weird results if the cable is sufficiently long and the VNA span is really wide (to get good resolution) that the maximum time is shorter than the time to propagate to the end of the cable and back.

Maybe you should have tried it first.? :-)

Mike - M0MLM

Today I blundered myself, I put shame on myself,
I disgraced myself, and I made a fool of myself
(with the help of a dictionary)
by trying to demonstrate a TDR measurement with the NANOVNA
to some radio amateur audience.

What a shit this is. To hell with it.


Using the latest NANOVNASAVER software the cable lenght window scaling
was about 4000 meters. Not possible to readout something meaningful.
No obvious way to change that. Spent a lot of time fiddling around.

My cable is 8,6 meters long, says the RIGEXPERT AA-230 ZOOM device.
And it does that on the press of a button.
And it displays a nice graphic.

Amateur crap versus professionals.

I am totally upset.

OE8UWW.

There is nothing wrong with being an Appliance Operator, we need them to keep the ham population large enough to support the ARRL and to prevent the corporate interests from lobbying to abolish Amateur Radio altogether so that they can take our spectrum and sell it back to us on a monthly subscription plan. I do get the impression sometimes that a lot of hams would gladly let go of the microwave spectrum if they could trade it for an extra few kilohertz added on to the 20 meter band.

I also note that 50 years ago having a ham license would put you at the top of the queue for most electronics and engineering jobs, but today most engineering hiring managers (and most engineering professors) have never heard of ham radio.

Dan N8FGV

Dear Bill,

I am also a retired doctor, who started playing with electronics in 1960, got my theory pass towards amateur licence age 14, and decided to keep radio/electronics/vintage computing as a hobby to keep me sane during the brutal years of my training to be a consultant, although I did pick up an electronics degree in 1972 along the way and network analyser was never heard of, let alone spectrum analyser or vna. I did see an analogue sampling oscilloscope once ( how many people onow shat one of those is).

However I do try to understand the complexities of this kit, avoid anything tooooo complex, and rant at impossible menus and apalling machine- human interfaces.

Nevertheless, I probably am an appliance operator who prefers an analogue scope to a digital one. Trying to use these little new toys is an exciting challenge however.

Steve L. G7PSZ

On 9/15/20 3:55 AM, Piero Tognolatti wrote:

Hi Colin,
Please, note that s2p file is in the "RI" format, your formula to compute SWR is not correct.?? "RI" means that one column (2nd column) shows the real part of S11, while the other column (3rd column) the imaginary part of S11.
Your formula is correct when the s2p file is in the "MA" format, which means that one column is the Magnitude of the S11, while the other column is the Angle of the S11. Let me recast your formula saying that SWR = (1+M) / (1-M)?? where M is the magnitude of S11, taken from the proper column of a s2p file in MA format.
At the moment I cannot run nanoVNA-saver, so I don't know if the exported s2p file are in MA or RI format.

regards
Piero, I0KPT

if it *is* in real/imaginary form, then

SWR = (SQRT(R^2+I^2)+1)/(SQRT(R^2+I^2-1)

Il 14/09/2020 22:51, Colin McDonald ha scritto:

With some more research I managed to create a formula to show swr in a SNP spread sheet.
SWR = (MA+1)/(1-MA)
Where MA (or RI) is the second colomn in the spread sheet.
for instance SWR = (B3+1)/(1-B3)

Come to think of it Bill, I AM an appliance operator. But I still don’t understand all the complexities of our microwave, TV, electric range, refrigerator. I have to refer often to the manuals (and they don’t write Nifty manuals for those items) to even operate some of their neat features. Would I go back to the appliances my mother grew up with? Heck no!

Like may daughter reminds us when she doesn’t do things the way we would do them, “It is what it is.”

Bottom line. Life is good even with its warts and moles. This is a wonderful hobby and it likely kept me away from the scent of cologne long enough to make a wise decision (married to the same woman for 47 years) about career and family. Luckily born a NERD with the “knack” as SolderSmoke Bill would say....

Back to having fun. Visiting the grands in Illinois and Thursday leave for Indiana to see more grands, one of whom is a future ham (a NERD like his Papa).

Dave K8WPE

David J. Wilcox K8WPE’s iPad

Hi,

I wonder where "dumbing down" and other insults come from. There are a lot of nice appliances available and they need appliance operators to get them on the air. I doubt there are two hams with the same exact approach to the hobby. Our diversity allows ham radio to survive and thrive.

I am pretty sure I have some things in my skillsets that most other people "can't touch". That does not make all those other people any less valuable. I notice other people have things in their toolboxes that I do not have. When we encourage other people and cooperate we are all better off. Let's lighten up with the ugly labels, please.

73,

Bill KU8H

bark less - wag more

Hi Colin,

Please, note that s2p file is in the "RI" format, your formula to compute SWR is not correct.?? "RI" means that one column (2nd column) shows the real part of S11, while the other column (3rd column) the imaginary part of S11.

Your formula is correct when the s2p file is in the "MA" format, which means that one column is the Magnitude of the S11, while the other column is the Angle of the S11. Let me recast your formula saying that SWR = (1+M) / (1-M)?? where M is the magnitude of S11, taken from the proper column of a s2p file in MA format.

At the moment I cannot run nanoVNA-saver, so I don't know if the exported s2p file are in MA or RI format.

regards

Piero, I0KPT

Is it “dumbing down” or increasing complexity? Back in 1960 when I was first licensed I was young and had a much quicker brain and felt I understood tubes, transistors, and RF circuitry sufficient enough to build and operate my Heathkit station. As time went on and Med school, family obligations, and career took over most of my life I honestly tried to keep up but if you want to understand and use much of what is available today you probably need an engineering degree or the equivalent. Sorry if I am now classed as an appliance operator. I would probably even flunk that title as there is so much more in each of my rigs that I don’t understand........ but I am still having fun and enjoy these varied .io groups. Call me what you will. This is supposed to be a hobby and a hobby is supposed to redirect your mind and heart away from the trials and tribulations of every day life. Ham radio still does that for me. Maybe we need new classes of license: Appliance Operator, Appliance Operator with CW, etc.

Dave K8WPE

David J. Wilcox K8WPE’s iPad

I am sorry you have had such a bad time with these little devices.

I have FOUR of them in different guises ( nano, V2, V2 with 3.2 screen, V2N) and they all work fine.

It needs a lot of patience to use them, partly because they are so small, and if it is without a case, it is also vulnerable to the environment as well. Also they need careful setting up compared to the MFJ kit.

I find the uses for these devices expands as you learn how to operate them. The “sister” spectrum analyser (Tinysa) is also very useful and in a case, but it still has to be understood to get the best out of it. It IS easier to use than a bench boatanchor one though.

Steve L. G7PSZ

Hi Alezis,

because of my weekend family activities, I just got here.
Please, if the problem is still current, visit my web page where you will find an answer to the solution.
I have v3.0 hardware from NanoVNA (it may be a July 2019 production) that I have converted to be much more capable in its current state. You can read the description here:
I hope you can solve the problem after reading the above.
You can also install the latest firmware.

73, Gyula HA3HZ
--
*** If you are not part of the solution, then you are the problem. ( ) ***