I set up my NanoVNA to measure SWR on my Butternut antenna.  I calibrated the VNA from 50KHz - 40MHz, O, S, L. Set up a Smith chart and a SWR graph.  Cal shows DTX on the left side. Since this antenna covers 8 bands, I set up the start and stop frequencies to cover each band, ie 14 - 14,5 MHz, start and stop, for the 20M band.  Likewise for the others.  On all the bands I get an SWR about 1.2 across the whole band, no matter which band I choose.  If I scan the entire ham band, I still get the 1.2 SWR!!   Either this is the best antenna that I have seen in my life, or I am doing something wrong here.  I really don't think that I did anything wrong setting up the VNA, so I'm betting on antenna problems. I did disconnect the coax at the Butternut antenna and shorted out the center conductor and the shield to eliminate static build up before applying the VNA to the Butternut's 75 ohm matching stub. I did measure the coax to the antenna using the VNA and a pulse generator and transform and velocity factor. Worked just fine!!!
Any input on this?
syd/ wt1v

You have calibrated over too wide of a frequency range. You need to cal
for EACH band. Given a finite number of points, a cal between 50 kHz and
40 MHz will not give enough resolution to measure each band accurately.
Calibrate each band alone and measure on that band only with the cal
dedicated to that band, alone.

Dave W0LEV

On Mon, Sep 26, 2022 at 11:32 PM Syd via groups.io <nhuq1=
yahoo.com@groups.io> wrote:

> I set up my NanoVNA to measure SWR on my Butternut antenna. I calibrated
> the VNA from 50KHz - 40MHz, O, S, L. Set up a Smith chart and a SWR graph.
> Cal shows DTX on the left side. Since this antenna covers 8 bands, I set up
> the start and stop frequencies to cover each band, ie 14 - 14,5 MHz, start
> and stop, for the 20M band. Likewise for the others. On all the bands I
> get an SWR about 1.2 across the whole band, no matter which band I choose.
> If I scan the entire ham band, I still get the 1.2 SWR!! Either this is
> the best antenna that I have seen in my life, or I am doing something wrong
> here. I really don't think that I did anything wrong setting up the VNA,
> so I'm betting on antenna problems. I did disconnect the coax at the
> Butternut antenna and shorted out the center conductor and the shield to
> eliminate static build up before applying the VNA to the Butternut's 75 ohm
> matching stub. I did measure the coax to the antenna using the VNA and a
> pulse generator and transform and velocity factor. Worked just fine!!!
> Any input on this?
> syd/ wt1v
>
>
>

--
*Dave - WØLEV*
*Just Let Darwin Work*

Not to hijack your thread, but I too am having questions about measuring a ham antenna SWR.  I calibrated my nVNA from 130MHz to 160MHz to cover the range around the ham 2-meter band and got a poor SWR figure (blue line) from 2.75 @ 144MHz DOWN to 1.64 @ 148MHz, with a Z_re peak about of 82 Ohms @ 147MHz.
To get a closer look I simply narrowed the stimulus sweep to 144MHz - 148MHz, without recalibration, and I got a very different curve.  Now the SWR goes from 1.8 @ 144 UP to 3.0 @ 148MHz, and the Z-re peak shifted to about 105 @ 145.4MHZ.
Why would narrowing the stimulus sweep change the values at specific frequencies so much?
SherpaDoug WA1UWP

Hi Douglas,

Always calibrate your VNA to the frequency range of interest for the best
results using the same fixture. Also, please make sure on your desktop PC
software you are using the most calibration points you can and, if you can
afford the time, average your sweeps.

The larger the frequency range of calibration the less accurate the
measurement will be for small frequency ranges.

If I am going to measure 144-148MHz I calibrate with at least 2,000 points
(2kHz per point) vs using a calibration of 50kHz to 900 MHz with 2,000
points (approximately .5MHz per point). At the higher calibration range I
would only have about 8 points measured and the VNA extrapolates with these
8 points over the entire 4 MHz wide band. As you can see the more
measuring points per MHz the more accurate the results.

Good luck!

Hank




On Thu, Sep 29, 2022 at 8:36 AM Douglas Butler <sherpadoug@comcast.net>
wrote:

> Not to hijack your thread, but I too am having questions about measuring a
> ham antenna SWR. I calibrated my nVNA from 130MHz to 160MHz to cover the
> range around the ham 2-meter band and got a poor SWR figure (blue line)
> from 2.75 @ 144MHz DOWN to 1.64 @ 148MHz, with a Z_re peak about of 82 Ohms
> @ 147MHz.
> To get a closer look I simply narrowed the stimulus sweep to 144MHz -
> 148MHz, without recalibration, and I got a very different curve. Now the
> SWR goes from 1.8 @ 144 UP to 3.0 @ 148MHz, and the Z-re peak shifted to
> about 105 @ 145.4MHZ.
> Why would narrowing the stimulus sweep change the values at specific
> frequencies so much?
> SherpaDoug WA1UWP
>
>
>

Agree!

On Thu, Sep 29, 2022 at 8:36 AM W0LEV <davearea51a@gmail.com> wrote:

> You have calibrated over too wide of a frequency range. You need to cal
> for EACH band. Given a finite number of points, a cal between 50 kHz and
> 40 MHz will not give enough resolution to measure each band accurately.
> Calibrate each band alone and measure on that band only with the cal
> dedicated to that band, alone.
>
> Dave W0LEV
>
> On Mon, Sep 26, 2022 at 11:32 PM Syd via groups.io <nhuq1=
> yahoo.com@groups.io> wrote:
>
>> I set up my NanoVNA to measure SWR on my Butternut antenna. I calibrated
>> the VNA from 50KHz - 40MHz, O, S, L. Set up a Smith chart and a SWR graph.
>> Cal shows DTX on the left side. Since this antenna covers 8 bands, I set up
>> the start and stop frequencies to cover each band, ie 14 - 14,5 MHz, start
>> and stop, for the 20M band. Likewise for the others. On all the bands I
>> get an SWR about 1.2 across the whole band, no matter which band I choose.
>> If I scan the entire ham band, I still get the 1.2 SWR!! Either this is
>> the best antenna that I have seen in my life, or I am doing something wrong
>> here. I really don't think that I did anything wrong setting up the VNA,
>> so I'm betting on antenna problems. I did disconnect the coax at the
>> Butternut antenna and shorted out the center conductor and the shield to
>> eliminate static build up before applying the VNA to the Butternut's 75 ohm
>> matching stub. I did measure the coax to the antenna using the VNA and a
>> pulse generator and transform and velocity factor. Worked just fine!!!
>> Any input on this?
>> syd/ wt1v
>>
>>
>
> --
> *Dave - WØLEV*
> *Just Let Darwin Work*
>
>
>
>

Hi Douglas,

The answer is in your pictures:-

First picture has Calibration applied, the Second picture has no Calibration
applied!!!

Cheers
Gary (G3TMG)

At 2000 points of calibration, over a 40 MHz range, that gives .02 MHz per point of calibration.  If I then select, for instance, to sweep the antenna from 3 - 4 MHz, to look at the 40 meter band, I should have 50 points shown on the screen which should be enough to show a dip in the SWR,  If I get the dip, then I can re calibrate the VNA to that particular band, 3 - 4 MHz in this case, to get a more accurate picture of the SWR! Sweeping each band after calibrating for the entire ham band doesn't seen like that bad an idea for looking at the antenna response on each of it's bands just to verify that there is some sort of dip. That does not seem that far fetched to me.  Yes going back and re calibrating for each band would be more accurate, but I just wanted to see if the Butternut was working before I started to investigate resonant problems! Also, as far as I know this unit calibrates 2000 points no matter the calibration range selected; therefore, the bigger the range, the less accuracy of the plot; hence, my calculations of the .02 MHz per point of calibration above.
syd / wt1v

Just a few comments from a total newbie to the NanoVNAV2.  First the question of the number of frequency increments in a scan. I can find remarkedly little reference to that point in the documentation.  What I can find is a reference to '201 steps maximum' for the NanoVNAV2 itself and then some comments relative to the companion PC based software of '..up to 2,000 points' and some comment regarding '...interpolation between data points...'

So my bewildered feeling at this point (I am going to ask for clarification from the community in a separate posting) is that the NanoVNAV2 itself only has a maximum number of frequency steps of 200.  And that, when using the PC based analysis program, you can get up to 2,000 steps.

So I would recommend you use the lower number for now.

One last point: regardless of all the above, it does look (strongly! in my opinion) that you have a problem with your antenna, it's feedline, etc.  You should absolutely see dips for the resonant frequencies. If you don't, something is wrong.

One more 'last point' - I would construct a simple LC network on the bench (say a series resonant LC network) with know values for the L and C and then sweep it.  You should be able to predict v more or less what you are should see. Good way to sneak up on things.

73
Bob Groh, WA2CKY

QUOTE: "So my bewildered feeling at this point (I am going to ask for
clarification from
the community in a separate posting) is that the NanoVNAV2
itself only has a
maximum number of frequency steps of 200. And that, when
using the PC
based analysis program, you can get up to 2,000 steps."

You are correct. The native storage capability is quite limited. Using
the PC-supported options frees that constraint by providing many more
points supported by the vastly expanded storage capabilities of the PC.

QUOTE: "One more 'last point' - I would construct a simple LC network on
the bench
(say a series resonant LC network) with know values for the L
and C and then
sweep it. You should be able to predict v more or less what
you are should
see. Good way to sneak up on things."

Known parameter bench tests are always a good idea if in doubt. They also
serve as excellent learning tools for the new owner of the NANOVNAs.

Dave - WØLEV

On Sun, Oct 16, 2022 at 5:01 PM Bob Groh wa2cky <bob.groh@gmail.com> wrote:

> Just a few comments from a total newbie to the NanoVNAV2. First the
> question of the number of frequency increments in a scan. I can find
> remarkedly little reference to that point in the documentation. What I can
> find is a reference to '201 steps maximum' for the NanoVNAV2 itself and
> then some comments relative to the companion PC based software of '..up to
> 2,000 points' and some comment regarding '...interpolation between data
> points...'
>
> So my bewildered feeling at this point (I am going to ask for
> clarification from the community in a separate posting) is that the
> NanoVNAV2 itself only has a maximum number of frequency steps of 200. And
> that, when using the PC based analysis program, you can get up to 2,000
> steps.
>
> So I would recommend you use the lower number for now.
>
> One last point: regardless of all the above, it does look (strongly! in my
> opinion) that you have a problem with your antenna, it's feedline, etc.
> You should absolutely see dips for the resonant frequencies. If you don't,
> something is wrong.
>
> One more 'last point' - I would construct a simple LC network on the bench
> (say a series resonant LC network) with know values for the L and C and
> then sweep it. You should be able to predict v more or less what you are
> should see. Good way to sneak up on things.
>
> 73
> Bob Groh, WA2CKY
>
>
>

--
*Dave - WØLEV*
*Just Let Darwin Work*

On 10/16/22 8:55 AM, Bob Groh wa2cky wrote:
> Just a few comments from a total newbie to the NanoVNAV2.  First the
> question of the number of frequency increments in a scan. I can find
> remarkedly little reference to that point in the documentation.  What I
> can find is a reference to '201 steps maximum' for the NanoVNAV2 itself
> and then some comments relative to the companion PC based software of
> '..up to 2,000 points' and some comment regarding '...interpolation
> between data points...'
>
> So my bewildered feeling at this point (I am going to ask for
> clarification from the community in a separate posting) is that the
> NanoVNAV2 itself only has a maximum number of frequency steps of 200.
> And that, when using the PC based analysis program, you can get up to
> 2,000 steps.


What the PC software does is command the NanoVNA to make multiple sweeps
over subsets of the range, each 101 or 201 steps, to build up the 2000
steps in the final range.

Also don't want to hijack the thread, but...
Did my first SWR test, of a HF random wire antenna, yesterday.  No calibration, just a sweep from 2 to 30MHz to get a ballpark figure.

What concerned me was: when my hand touched the metal case, the SWR reading dropped.

I'm *guessing* I have a bad counterpoise connection, and my body is making a surrogate counterpoise down at the nanvna.
Has anyone else had a similar experience?

Alan - VK5OUP

This is common and expected without a proper counterpoise. For example,
when making measurements at the antenna feedpoint of a dipole, one must
completely isolate the VNA from anything and everything. No USB
connections.....no charger connections......no "flesh" connections.....well
isolated from soil surface....well isolated from anything even marginally
conductive.

Dave - WØLEV

On Thu, Nov 10, 2022 at 4:31 AM Alan Campbell <soupie62@gmail.com> wrote:

> Also don't want to hijack the thread, but...
> Did my first SWR test, of a HF random wire antenna, yesterday. No
> calibration, just a sweep from 2 to 30MHz to get a ballpark figure.
>
> What concerned me was: when my hand touched the metal case, the SWR
> reading dropped.
>
> I'm *guessing* I have a bad counterpoise connection, and my body is
> making a surrogate counterpoise down at the nanvna.
> Has anyone else had a similar experience?
>
> Alan - VK5OUP
>
>
>

--
*Dave - WØLEV*
*Just Let Darwin Work*