开云体育

On Sat, Apr 26, 2025 at 10:39 AM, Dean W8ZF wrote:

because MAYBE it depends on the H4's cal state

From the nanovna-saver github readme ():
Quote:
Calibration

Before using NanoVNA-Saver, please ensure that the device itself is in a reasonable calibration state.

A calibration of both ports across the entire frequency span, saved to save slot 0, is sufficient. If the NanoVNA is completely uncalibrated, its readings may be outside the range accepted by the application.

-------------
And from the code itself, the following code fragment shows that nanovna-saver uses the "scan" command, rather than the "scanraw" command, to get its data from nanovna. The "scan" command returns S11/S12 data with nanovna internal calibration (if any) applied.

def setSweep(self, start, stop):
self.start = start
self.stop = stop
list(self.exec_command(f"scan {start} {stop} {self.datapoints}"))
-------------

So nanovna-saver indeed does depend on the calibration state of the nanovna; they are not independent.
Q.E.D.

If you want independent calibration, use nanovna-app instead. It is a quite different application with its own learning curve, but it has specific settings to use the nanovna internal or nanovna-app calibration. I use both apps, they both have excellent features.

Stan KC7XE

And thank you, Stan, for your feedback!
73,
Dean W8ZF

Thanks Dave,
So if the H4 internal cal is *independent* of the nanovna-saver cal, exactly what are you clearing with nanovna-saver's "clear all cals"and when would that be necessary?
I ALWAYS either calibrate prior to measurements (and upon any sweep range or source power change) or I alternatively load a saved calibration file generated by nanovna-saver. Doesn't using the calibration assistant or loading the cal file "clear all cals"?

I understand the principles and use of a VNA. I'm a Senior Principal RF engineer and have used professional VNAs for 45 years. But I am new to the nanovna VNA implementation and the interactions of the H4 with software interfaces. What would be the source of information on whether nanovna-saver uses raw data from the nanovna as part of its calibration, nor not? I can always use the internal calibrate routine in the H4 for mobile use, it looks like that's necessary (since the nanovna-saver cal data is independent). But I am still not convinced that I can trust nanovna-saver's calibration (because MAYBE it depends on the H4's cal state).

Thanks and 73,
Dean W8ZF

No. They are independent.

Dave - W?LEV

On Sat, Apr 26, 2025 at 5:00?PM Stan Dye via groups.io <standye=
[email protected]> wrote:

Dave, correct information, but you did not address his real question: Is
the nanovna-saver 0.4.0 calibration dependent on the nanovna internal cal?

I was taught (back a couple of years ago) that nanovna-saver reads the
calibrated data from the nanovna. So on nanovna, you should do a
wide-range calibration, and use that cal whenever you use nanovna-saver
calibrations - both during the nanovna-saver calibration, and for all
measurements with nanovna-saver. Alternately, you could reset/disable the
nanovna calibration whenever you use nanovna-saver, but sometimes that
caused problems due to an out-of-range raw data value from nanovna. Back
at that time, I verified that this was correct.

Has nanovna-saver been changed to read the raw data and deal with the
out-of-range problem? Not that I am aware of. [But then I mostly use
nanovna-app nowadays.]

And Dean, in answer to your question, if you always use the same nanovna
cal when using nanovna-saver, no nanovna change is necessary when you
change ranges, etc. and calibrate in nanovna-saver. This is because the
nanovna-saver calibration compensates for whatever it sees in the data from
nanovna. Ideally, the nanovna calibration will have a wide frequency
range which is a superset of the frequency ranges that you will use in
nanovna-saver.

--

*Dave - W?LEV*


--
Dave - W?LEV

Dave, correct information, but you did not address his real question: Is the nanovna-saver 0.4.0 calibration dependent on the nanovna internal cal?

I was taught (back a couple of years ago) that nanovna-saver reads the calibrated data from the nanovna. So on nanovna, you should do a wide-range calibration, and use that cal whenever you use nanovna-saver calibrations - both during the nanovna-saver calibration, and for all measurements with nanovna-saver. Alternately, you could reset/disable the nanovna calibration whenever you use nanovna-saver, but sometimes that caused problems due to an out-of-range raw data value from nanovna. Back at that time, I verified that this was correct.

Has nanovna-saver been changed to read the raw data and deal with the out-of-range problem? Not that I am aware of. [But then I mostly use nanovna-app nowadays.]

And Dean, in answer to your question, if you always use the same nanovna cal when using nanovna-saver, no nanovna change is necessary when you change ranges, etc. and calibrate in nanovna-saver. This is because the nanovna-saver calibration compensates for whatever it sees in the data from nanovna. Ideally, the nanovna calibration will have a wide frequency range which is a superset of the frequency ranges that you will use in nanovna-saver.

QUOTE: ........ but what happens when the sweep range is different for
nanovna-saver than the H4's stored cal? Are results invalid, or are they
interpolated (assuming nanovna-saver is sweeping in a sub-range of the H4's
calibrated range)?

When you do a cal in SAVER, there is an option to clear all previous cals.
This SHOULD BE EXERCISED to assure a clean and meaningful new cal.

When you change the sweep width, it is always an excellent idea to do a new
cal dedicated to the new intended frequency sweep range. If the "new"
range is greater than the previous cal range, the instrument will use the
cal but interpolate between cal points. This potentially will introduce
extrapolation errors which could have been avoided by doing a dedicated new
cal over the "new" intended frequency sweep range. When changing sweep
ranges it's ALWAYS a good practice to run an additional cal dedicated to
the new range. This also applies to "professional" (and very expensive)
VNAs.

Dave - W?LEV

On Sat, Apr 26, 2025 at 3:59?PM Dean W8ZF via groups.io <dwfred=
[email protected]> wrote:

Thanks for the replies. There seems to be some disagreement on #2.
Appparently for some versions (maybe all), the cal in nanavna-saver IS
dependent on the H4 being in a calibrated state before the nanovna-saver
cal is performed. I can do the H4 internal cal, no problem, but what
happens when the sweep range is different for nanovna-saver than the H4's
stored cal? Are results invalid, or are they interpolated (assuming
nanovna-saver is sweeping in a sub-range of the H4's calibrated range)?

--

*Dave - W?LEV*


--
Dave - W?LEV

Thanks for the replies. There seems to be some disagreement on #2. Appparently for some versions (maybe all), the cal in nanavna-saver IS dependent on the H4 being in a calibrated state before the nanovna-saver cal is performed. I can do the H4 internal cal, no problem, but what happens when the sweep range is different for nanovna-saver than the H4's stored cal? Are results invalid, or are they interpolated (assuming nanovna-saver is sweeping in a sub-range of the H4's calibrated range)?

QUOTE: 1) What is the persistence of calibration done through
nanovna-saver? I.e., does it persist if the USB cord is disconnected and
the H4 is taken for mobile measurements?

Any cal done within SAVER stays with SAVER. Any cal done on the native
NANOVNA stays with the NANOVNA. They do not cross pathes.
***

2) Is the calibration done by nanovna-saver dependent on the calibration
done by the H4's internal cal? For instance, say memory #0 on the H4 has a
bad calibration stored. Since this is the default used by the H4 when it
boots up, will it corrupt the calibration done in nanovna-saver. Does
nanovna-saver depend on the calibration state of the H4 to be correct?

The cals done on the native NANOVNA and within SAVER are independent of
eachother.
*

3) Is is possible to store the calibration done in nanovna-saver to one of
the memory states in the H4? Storing it in memory #0 would allow for a
calibration to be persistent through a power cycle.

The cals between the native NANOVNA and within SAVER are independent.
\

It's all like what happens in Vegas stays in Vegas.......

Dave - W?LEV


On Sat, Apr 26, 2025 at 2:49?PM Dean W8ZF via groups.io <dwfred=
[email protected]> wrote:

I use nanovna-saver with my H4 because it aggregates several useful
functions for operating, storing, printing, and exporting measurements. I
am confused about the calibration done from nanovna-saver. I know I can
load and save calibrations on my PC as files. This works and I use it all
the time. I wanted to take the H4 out to the yard and do measurements at
the antenna after I calibrated with nanovna-saver. It seems that it did
not retain the calibration. I don't remember if I had turned the H4 off,
then back on, but the results were different when I then brought out my
laptop and loaded my calibration file. Everything seemed to measuring
correctly, after that. So I have some questions;
1) What is the persistence of calibration done through nanovna-saver?
I.e., does it persist if the USB cord is disconnected and the H4 is taken
for mobile measurements?
2) Is the calibration done by nanovna-saver dependent on the calibration
done by the H4's internal cal? For instance, say memory #0 on the H4 has a
bad calibration stored. Since this is the default used by the H4 when it
boots up, will it corrupt the calibration done in nanovna-saver. Does
nanovna-saver depend on the calibration state of the H4 to be correct?
3) Is is possible to store the calibration done in nanovna-saver to one
of the memory states in the H4? Storing it in memory #0 would allow for a
calibration to be persistent through a power cycle.

Thanks and 73,
Dean W8ZF

--

*Dave - W?LEV*


--
Dave - W?LEV

#2 is correct.
The calibration done in nanovna-saver is solely in the computer. It can't be saved to a slot in the nanovna, partly because -saver has more sweep-point options. You need to do the calibration on nanovna separately.

Also, it is important to note that (in ver. 0.4.0 and maybe others) the calibration done in nanovna-saver runs on top of the active calibration in the nanovna. [Some later version which I am not aware of may have a selection to use raw nanovna data.]

So it is important to save a basic broad calibration in nanovna slot 0, or a slot you will recall before each nanovna-saver calibration and use. Alternatively always reset or disable nanovna calibration while using nanovna-saver. [But sometimes -saver will get an error with the latter approach, due to out-of -range data from nanovna.]

I use nanovna-saver with my H4 because it aggregates several useful functions for operating, storing, printing, and exporting measurements. I am confused about the calibration done from nanovna-saver. I know I can load and save calibrations on my PC as files. This works and I use it all the time. I wanted to take the H4 out to the yard and do measurements at the antenna after I calibrated with nanovna-saver. It seems that it did not retain the calibration. I don't remember if I had turned the H4 off, then back on, but the results were different when I then brought out my laptop and loaded my calibration file. Everything seemed to measuring correctly, after that. So I have some questions;
1) What is the persistence of calibration done through nanovna-saver? I.e., does it persist if the USB cord is disconnected and the H4 is taken for mobile measurements?
2) Is the calibration done by nanovna-saver dependent on the calibration done by the H4's internal cal? For instance, say memory #0 on the H4 has a bad calibration stored. Since this is the default used by the H4 when it boots up, will it corrupt the calibration done in nanovna-saver. Does nanovna-saver depend on the calibration state of the H4 to be correct?
3) Is is possible to store the calibration done in nanovna-saver to one of the memory states in the H4? Storing it in memory #0 would allow for a calibration to be persistent through a power cycle.

Thanks and 73,
Dean W8ZF

On Fri, Apr 25, 2025 at 01:30 PM, KK7OYV wrote:

New here. Any tricks to make calibration easier? It is a real hassle as it is.

Surprised you said that. Only takes about 2 minutes for the whole procedure. And if you don't care about S21 measurements you can skip the last two steps and only do SOL (short-open-load)

Roger

If you use SAVER on a PC or laptop, you can store your verious cals on
those and access them later for use. Calibrate once for each frequency
range and setup.....store on a "big" PC........recall for use.

Other than that, yes, you should cal. for every change in frequency range.
That's only three standards! Of course, you must first remember to clear
any pre-existing cals and store the new cal when done.

Dave - W?LEV

On Fri, Apr 25, 2025 at 8:30?PM KK7OYV via groups.io <kk7oyv=
[email protected]> wrote:

New here. Any tricks to make calibration easier? It is a real hassle as it
is.

--

*Dave - W?LEV*


--
Dave - W?LEV

New here. Any tricks to make calibration easier? It is a real hassle as it is.

Re: rain gutter antenna

I also have one rain gutter antenna it's about 50 ft long and 8 ft above
ground level. For one 40-meter freq i use a 1:1 balun (indoor 10-meter-long
radial) plus a matching network i designed using the nanoVNA. Low power,
about 5W, talking on CW across town but haven't tried any distant contacts.

The same rain gutter and radial i've used on 20 meter WSPR using a 9:1
balun with no add'l matching network. I was heard worldwide on 20m WSPR,
when conditions were good, but that's not a 2-way contact.

Anyway i have better antennas than that, but it's fun to play with.
Apartment complex, so i have an attic dipole and an invisible wire dipole
on the roof (13 ft AGL) that i was allowed to have because the wires are
tucked under the shingles and are almost impossible to see.

Best 73 de WN1Z
Orrin


El vie, 25 abr 2025 a la(s) 9:00?a.m., W0LEV via groups.io (davearea51a=
[email protected]) escribió:

Out of a major forest fire, insurance put us up in a large apartment
complex for 2 years. I chose the 3d floor - the top - for RF reasons. Of
course, the apartment complex came with its manditory set of rules:
Absolutely no antennas of any kind. However, being on the 3d floor, we had
quite an assembly of rain gutters on the external surface of the building.
The light went on......

I used the small metal railing on the miniature patio as a counterpoise. I
snaked two wires through a drain hole at the bottom of the sliding door to
the miniature patio. One wire connected to the coax braid and the other
wire connected to the center conductor. The two wires were twisted to form
a "sort'a" transmission line. Of course, the briad wire connected to the
railing (after scraping a small amount of paint). The center conductor
wire connected to one of the downspouts (after scraping a bit of paint).
Once inside, I made an "ugly balun" by winding extra coax around a 3"
diameter cardboard tubing. I vaguely remember around 15 turns of coax on
the tube. From there, the coax was routed to the loft and fed an antenna
"tuner".

Well, it worked OK. But the RFI generated within the apartment complex was
so bad, I could detect it on a crystal set using the rain gutter antenna
directly connected to the crystal radio. Yes, I made a few contacts and
had excellent reports, but the RFI severely compromised reception.

Yes, rain gutters do work as makeshift - and "invisible" and unsuspecting -
antennas. However, one morning I got the bug to attempt 75-meters with the
antenna "tuner". That was a mistake. At the low power level of 20-watts,
I set off the fire alarm throughout the entire complex (about 2 city blocks
in area). Would I do it again? Absolutely!!!

Fortunately, we no longer reside there!!!

Dave - W?LEV

On Fri, Apr 25, 2025 at 3:33?PM Barry K3EUI via groups.io <k3euibarry=
[email protected]> wrote:

*
*

*How good can a RAIN GUTTER be on HF bands?*

*
*

*Since moving to a "retirement community" last year I have been
searching for a reasonably good HF antenna.*

*I have tried a few 10-20 ft verticals (not well hidden) and a 140 ft
long wire end-fed wire (which everyone can see).*

*
*

*So how about a commercial aluminum (painted) rain gutter.*

*I feed it with 50 ft RG8X into a Palomar Engineering 9:1 UNUN
with two counterpoises: 15 ft and 30 ft.*

*I have a wide-range tuner in the shack, so now SWR issues.
*

*Well, it is "hidden" and it does tune up on 160m, 40m, 30m, 17m, and
10 meters with SWR < 2.5:1*

*but that may mean it is just "lossy" (the UNUN).*

*
*

*I'll give it a try this weekend with Fldigi, running no more than 25
watts on CW, PSK, and Olivia/Thor modes.*

*I'm really curious about 160m at night - I have never operated on that
band.*

*Are there "watering holes" (freq) where different sound card modes hang
out?*

*Seems like everything is FT8 these days.
*

*10m should be easy, as it is both vertical and horizontal polarization.
*

*
*

*Here are Nano VNA SWR and RETURN LOSS graphs.*

*I plot RETURN LOSS as a negative number to have the same shape as the
SWR curve.
*

*I did one run from 1.8 to 8 MHz, and the other run from 8 to 30 MHz.
*

*I'm excited about this new antenna. No one can "see it".*

*But how well will it work?
*

*de k3eui barry*

--

*Dave - W?LEV*


--
Dave - W?LEV

Out of a major forest fire, insurance put us up in a large apartment
complex for 2 years. I chose the 3d floor - the top - for RF reasons. Of
course, the apartment complex came with its manditory set of rules:
Absolutely no antennas of any kind. However, being on the 3d floor, we had
quite an assembly of rain gutters on the external surface of the building.
The light went on......

I used the small metal railing on the miniature patio as a counterpoise. I
snaked two wires through a drain hole at the bottom of the sliding door to
the miniature patio. One wire connected to the coax braid and the other
wire connected to the center conductor. The two wires were twisted to form
a "sort'a" transmission line. Of course, the briad wire connected to the
railing (after scraping a small amount of paint). The center conductor
wire connected to one of the downspouts (after scraping a bit of paint).
Once inside, I made an "ugly balun" by winding extra coax around a 3"
diameter cardboard tubing. I vaguely remember around 15 turns of coax on
the tube. From there, the coax was routed to the loft and fed an antenna
"tuner".

Well, it worked OK. But the RFI generated within the apartment complex was
so bad, I could detect it on a crystal set using the rain gutter antenna
directly connected to the crystal radio. Yes, I made a few contacts and
had excellent reports, but the RFI severely compromised reception.

Yes, rain gutters do work as makeshift - and "invisible" and unsuspecting -
antennas. However, one morning I got the bug to attempt 75-meters with the
antenna "tuner". That was a mistake. At the low power level of 20-watts,
I set off the fire alarm throughout the entire complex (about 2 city blocks
in area). Would I do it again? Absolutely!!!

Fortunately, we no longer reside there!!!

Dave - W?LEV

On Fri, Apr 25, 2025 at 3:33?PM Barry K3EUI via groups.io <k3euibarry=
[email protected]> wrote:

*
*

*How good can a RAIN GUTTER be on HF bands?*

*
*

*Since moving to a "retirement community" last year I have been
searching for a reasonably good HF antenna.*

*I have tried a few 10-20 ft verticals (not well hidden) and a 140 ft
long wire end-fed wire (which everyone can see).*

*
*

*So how about a commercial aluminum (painted) rain gutter.*

*I feed it with 50 ft RG8X into a Palomar Engineering 9:1 UNUN
with two counterpoises: 15 ft and 30 ft.*

*I have a wide-range tuner in the shack, so now SWR issues.
*

*Well, it is "hidden" and it does tune up on 160m, 40m, 30m, 17m, and
10 meters with SWR < 2.5:1*

*but that may mean it is just "lossy" (the UNUN).*

*
*

*I'll give it a try this weekend with Fldigi, running no more than 25
watts on CW, PSK, and Olivia/Thor modes.*

*I'm really curious about 160m at night - I have never operated on that
band.*

*Are there "watering holes" (freq) where different sound card modes hang
out?*

*Seems like everything is FT8 these days.
*

*10m should be easy, as it is both vertical and horizontal polarization.
*

*
*

*Here are Nano VNA SWR and RETURN LOSS graphs.*

*I plot RETURN LOSS as a negative number to have the same shape as the
SWR curve.
*

*I did one run from 1.8 to 8 MHz, and the other run from 8 to 30 MHz.
*

*I'm excited about this new antenna. No one can "see it".*

*But how well will it work?
*

*de k3eui barry*

--

*Dave - W?LEV*


--
Dave - W?LEV

*
*

*How good can a RAIN GUTTER be on HF bands?*

*
*

*Since moving to a "retirement community" last year I have been searching for a reasonably good HF antenna.*

*I have tried a few 10-20 ft verticals (not well hidden) and a 140 ft long wire end-fed wire (which everyone can see).*

*
*

*So how about a commercial aluminum (painted) rain gutter.*

*I feed it with 50 ft RG8X?? into a? Palomar Engineering? 9:1? UNUN? with two counterpoises: 15 ft and 30 ft.*

*I have a wide-range tuner in the shack, so now SWR issues.
*

*Well, it is "hidden" and it does tune up? on 160m, 40m, 30m, 17m, and 10 meters with SWR? < 2.5:1*

*but that may mean it is just "lossy" (the UNUN).*

*
*

*I'll give it a try this weekend with Fldigi, running no more than 25 watts on CW, PSK, and Olivia/Thor modes.*

*I'm really curious about 160m at night - I have never operated on that band.*

*Are there "watering holes" (freq) where different sound card modes hang out?*

*Seems like everything is FT8 these days.
*

*10m should be easy, as it is both vertical and horizontal polarization.
*

*
*

*Here are ? Nano VNA??? SWR?? and ? RETURN LOSS graphs.*

*I plot RETURN LOSS as a negative number to have the same shape as the SWR curve.
*

*I did one run from 1.8 to 8 MHz, and the other run from? 8 to 30 MHz.
*

*I'm excited about this new antenna.? No one can "see it".*

*But how well will it work?
*

*de k3eui?? barry*

On Wed, Apr 23, 2025 at 06:31 PM, Clyde Spencer wrote:

This can be improved by adding a small Common Mode Choke at the VNA end to
help isolate the coax from the case of the device.

Clyde, in HTs used for their purpose, which is portable communications, the body of the HT, and even the body of the operator are all part of the antenna. If you mount a rubber duckie directly on a NanoVNA, possibly on a metal bracket, and hold it in your hand, you emulate this situation, and get values close to the real thing. But if you isolate the rubber duckie from the instrument by means of a common mode choke, you will create a very different situation, and totally mess up the measurement!

A rubber duckie at the end of a coax cable will behave totally different than when mounted on an HT.

Manfred

This can be improved by adding a small Common Mode Choke at the VNA end to
help isolate the coax from the case of the device. The frequency being
measured also has an impact.

*Clyde K. Spencer*



On Wed, Apr 23, 2025 at 1:28?PM Kenneth Roberts via groups.io <kenr313=
[email protected]> wrote:

Yes, I have noted that the SWR changes when holding the VNA vs. not
holding it, and touching the SMA connector vs. not touching it. The
readings generally are better when touching vs. not touching.

Ken, W4KRR


On 4/23/2025 12:20 PM, Manfred Mornhinweg via groups.io wrote:

I have a whole load of rubber duck antennas that I'm checking so see at
what range they should work best.

Ken,
the behavior of rubber duckies depends significantly on the radio they

are attached to. So, to get reasonably accurate results when measuring them
with the nanoVNA, I suggest that you build a metal bracket about the size
of a typical HT, covered with some insulation to mimmick the plastic body
of an HT, mount the nanoVNA in that bracket, its measuring port bonded to
the bracket, and then hold that contraption in your hand like you would
hold an HT, while making the measurements.

You could even use a plastic box the size of an HT, covered with

aluminium foil on the inside, properly bonded to the nanoVNA's port.

Of course you need to run the nanoVNA without an USB cable connected.

You will see that depending on how you grab that test contraption, the

measured values will shift somewhat. If you mount the antenna directly to
the nanoVNA, they will shift more, since the nanoVNA is a lot smaller than
an HT, at least the 2.8" screen version. I haven't compared the size of the
H4 version to that of an HT. It should be closer, but maybe not close
enough.

Perhaps someone else can comment on how accurate the measurements come

out when mounting a rubber duckie directly on an H4?

Manfred

Yes, I have noted that the SWR changes when holding the VNA vs. not holding it, and touching the SMA connector vs. not touching it. The readings generally are better when touching vs. not touching.

Ken, W4KRR

I have a whole load of rubber duck antennas that I'm checking so see at
what range they should work best.

Ken,
the behavior of rubber duckies depends significantly on the radio they are attached to. So, to get reasonably accurate results when measuring them with the nanoVNA, I suggest that you build a metal bracket about the size of a typical HT, covered with some insulation to mimmick the plastic body of an HT, mount the nanoVNA in that bracket, its measuring port bonded to the bracket, and then hold that contraption in your hand like you would hold an HT, while making the measurements.

You could even use a plastic box the size of an HT, covered with aluminium foil on the inside, properly bonded to the nanoVNA's port.

Of course you need to run the nanoVNA without an USB cable connected.

You will see that depending on how you grab that test contraption, the measured values will shift somewhat. If you mount the antenna directly to the nanoVNA, they will shift more, since the nanoVNA is a lot smaller than an HT, at least the 2.8" screen version. I haven't compared the size of the H4 version to that of an HT. It should be closer, but maybe not close enough.

Perhaps someone else can comment on how accurate the measurements come out when mounting a rubber duckie directly on an H4?

Manfred