开云体育

Hi, I got one one of these a few days ago and what a great piece of kit for the price, I have grasped the basics of it with and getting on great, I tried the PC software NanoVNA Saver by Rune Broberg, which is also working great but are there any user manual or help files for the PC software. Thanks, Johnny GM7LSI

Something I have mentioned on other groups is in order here.

For sake of argument (and real life), consider only two types of SWR meters:
1) A meter with multiple power levels selected by switches.
Assume these are 20, 200, and 2000 Watts. These are usually the least
expensive
measurement devices, and, therefore, more prevalent among
the amateur community.
2) A good cross needle meter with the same ranges.

For #1, if the power incident from the transmitter is at the rated
calibration levels, i. e. 20, 200 and 2000 Watts (above legal power), the
meter will read correctly within its tolerance range (BTW, the Bird is
rated at 5% of full scale). If your transmitter puts out 100 watts and you
have the 200 watt scale selected, the SWR will read in error. If you run a
legal limit amplifier at 1500 watts output and you have selected the 2000
watt scale, the meter likely will read SWR incorrectly. This is due to not
being able to *normalize* the full-scale meter reading to the incident
power. If you don't understand normalization, look it up. Have a read
of: <> under the Mathematics
and Statistics subheading. Virtually all the old SWR meters (I had the
Knight unit some 50 years ago) required normalization to full scale in the
forward position for what ever incident power was fed through the sampled
line. They did not suffer the inaccuracies of today's meters that do no
allow for normalization to incident power. However, they did not pretend
to measure actual forward and reverse *power*.

For #2, the case of the previous paragraph is not so. Being a cross needle
meter, the intersection of the needles, representing forward and reverse
power, so-to-speak, are self-normalizing and will always read correctly
from the meter face within the tolerance of the meter.

Take the Bird (or equivalent - NOT MFJ, please) with a 2.5 kW head and
incident power of 1000 watts. 5% of 2.5 kW is +/-125 watts - the rated
tolerance of the Bird line of watt meters. That means that your 1000 watts
could read 1000 +/- 125 watts or 875 to 1125 watts, all within the rated
full scale tolerance of 5%. But maybe we have a peak reading meter on our
amplifier that shows 1001 watts for either speech or CW key-down. Which of
the three values are correct, 875 watts, 1125 watts, or 1001 watts? All
are correct within the tolerance of the measurement instrument.

Dave - W?LEV

--

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

Several reasons:

The SWR meter is not well calibrated. Simple diode detectors
have error and that varies with power.

Transmitter has harmonic content, that tends to increase SWR
as the harmonics may be outside the band the antennas is matched for.

The antenna system has RF on the shield. It can be due to imbalance,
mismatch at the radiating element end of the coax. In the case of the
transmitter the cable (coax) is in a area of high field strength and there
are induced currents on the shield that would be hard to see at much
lower power.

If you see different behavior from the VNA with USB power or longer
coax ( holding it vs laying on table) its a safe bet the antenna system
has issues.

Allison
-----------------
I do not accept private email due to forum scraping groups.io

Its a way to permit using attenuation where you need it
and allow for gain as needed.

I can be used with a different reference such as 75 ohms (any actually)
for systems that are not 50 ohms to make measurements.

An example using that bridge is measuring a radios input impedance
where a very small signal may over load the radio.

If you put a attenuator on port 0 (s11) the return is also attenuated
and the results are hard to discern.

With the bridge you can attenuate the output and see the returned
signal from the radio input using port2 (s21) and if the signal is
weak one can add an amplifier (suitable type) at the input to PORT2.


There are other measurements that may require that as well
for example at higher powers.

The basic instrument is very flexible but the basis of it is a
RF source that has a matching detector that can present
phase and amplitude information.

-----------------
I do not accept private email due to forum scraping groups.io

Hi Ron,


I first had a similar experience to your's:

On the NanoVNA calibrated up to 900 MHz, my broadbanded W3DZZ on 80m first
showed much "better SWR" than believable.
After the below described calibration to only the 300 KHz span of the 80m
band , the NanoVNA now shows very believable
(near the rig measured SWR=1:1.75 at 3.65 MHz) SWR now.

For an extra comparison I also used my well proven AA-600 Rigexpert. It is
close to the narrow band NanoVNA.

Reasons: (See jbrusgrove's answer that just came in: He sais the same.)

1. The NanoVNA calibration should be made to just cover the band limits of
the band (in my case 3.5 ... 3.8 MHz EU 80m band).
Do not use the general coverage like up to 900 MHz. Why? Each cal point then
is 9 MHz apart from the next, and interpolation gets extremely bad.
Larry's (W0QE) #89 video is right on that topic, showing that close
frequency point distance between the 101 points is important for good
interpolation:

Here it is:

So don't let the 101 points cover unnecessary out-of-interest parts.


2. The remaining (relatively small) difference imho mainly may have this
reason:

The voltages applied are different: A 100 Watt TX will go way beyond the SWR
meter's diode threshold voltage.
So the percentage of the diode caused error is small and even the low
reflected voltage is measured well enough.

The NanoVNA works with very small power. I am not certain, what actually it
is, but it is so much smaller.
I am also not certain about the internals of the NanoVNA, but somewhere some
threshold must play a role, too, I assume.
Especially when measuring low SWR, the backward voltage then (percentage
wise) is very much subject to diode error.


3. Besides the above, I think some of the other points mentioned already in
this thread are likely, too.

Except: I cannot confirm that changing the calibration plane by using an
extra UHF plug cal set would make any realizable difference on 80m
measuring:
You can save that effort: We talk about maybe 4 cm for the plug difference,
but we have Lambda = 80m so that Lambda/2000 really is not significant here.
All other influences mentioned in the thread should be larger than this
particular one.

The measures were taken at different times:

The broadbanding tuner settings were roughly, but not exactly the same.
Also wheather (ground conductivity) changes results, and as wind goes,
things change, too.

In spite of that, I got at 3.65 MHz:

At the TS-930 (without ATU) SWR meter needle: SWR = 1:1.75
At the AA-600: SWR = 1:1.7
At the (narrow cal) NanoVNA: SWR = 1:1.71

But at the (unnecessarily wide cal) NanoVNA: SWR = 1:1.18 (!!!)

That solved your problem, I think.


73, Hans
DJ7BA



-----Ursprüngliche Nachricht-----
Von: [email protected] <[email protected]> Im Auftrag von Ron -
An Old Ham in Utah
Gesendet: Sonntag, 17. November 2019 03:04
An: [email protected]
Betreff: Re: [nanovna-users] SWR...Nano versus Transmitting

The band that the primary problem is 40 meters; my xmtr is running on a
conventional power supply, not battery powered.

Hi YO2NAA -

what measurements could be improved, or what additional measurements can be
done if using an external RF bridge, like this one:

Funny that you show a blue one; if not directly from transverters-store.com,
then probably a bad clone. Specifically,
mine from eBay had paired 100 Ohm resistors to obtain 50 Ohms,
but one of each was not connected.

I've read a lot of messages from the group and some wiki files, I've seen
the bridge mentioned but I still can't figure out if it would be an useful
addition to NanoVNA.
I suspect using the bridge would improve measurement of high impedance
if using a reference higher than 50 ohm.

Those unconnected 100 Ohm resistors make this work better for higher impedances,
provided that a higher reference value is also employed
and coax wiring is corrected.
/g/nanovna-users/message/7205

Any info or examples are welcomed.

One challenge for use with nanoVNA is that I don't know any software
that directly supports calibrated S11 measurement by CH1,
which is wanted when using an external bridge.

Hi all,

what measurements could be improved, or what additional measurements can be
done if using an external RF bridge, like this one:
[image: image.png]

I've read a lot of messages from the group and some wiki files, I've seen
the bridge mentioned but I still can't figure out if it would be an useful
addition to NanoVNA.
I suspect using the bridge would improve measurement of high impedance if
using a reference higher than 50 ohm.

Any info or examples are welcomed.

Thank you,
73 YO2NAA

For best results always calibrate the same frequency range as you will be measuring.
With 900MHz range and only 101 data points the nanoVNA will interpolate the calibration data leading to relevant errors, SWR measurement is very susceptible to this.
--
NanoVNA Wiki: /g/nanovna-users/wiki/home
NanoVNA Files: /g/nanovna-users/files
Erik, PD0EK

To further communicate the point made by Paul, W1IP, the following may be
of some enlightenment.

If I take 101 points, the max for the NANOVNA and spread the evenly of the
max frequency range, that yields only 101 / (900 - 0.05) = 0.11 points per
MHz. You don't even cover the whole 75/80 meter band with one point!!!
You get only 0.1 points for the whole half MHz!! Of course, it will read
the same over the whole band.

Now, calibrate over 3.5 to 4 MHz: 101 / (4.0 - 3.5) = 202 points per MHz.
Of course, you only use the max of 101 points which covers the whole 75/80
meter band with good resolution. Now you have all 101 points devoted to
reflecting the measurement of the entire band - MUCH more accurate.

Please, before badding the instrument, read and digest the tutorials! No,
this isn't an HP 8753C, but it's also not your typical MFJ or Comet piece
of gear. Learn how to use a VNA. Don't feel bad as most electrical
engineers (even at the PhD level) don't understand how to properly use a
VNA or crawl around on a Smith Chart. Believe me as I worked with one at
HP who had absolutely no idea what an S-Parameter represented or what
return loss indicated. You'll never go back to MFJ!

Dave - W?LEV

--

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

Paul

Are you calibrated for and looking wideband 50 kHz - 900 MHz or calibrated and looking narrowband 3.5 - 4.0 MHz? Confining the 101 data points to the band of interest yields the most detail.

Jay W1VD

I'm with you Ron, this thing is not right. It show my 80m dipole as having a good match across the entire 75/80 band, lol. It is cut to favor about 3.9, and three analyzers (Comet, MFJ and Rigexpert) along with meter in my Expert linear all agree! But they must all be wrong. I did not expect much from a $50 unit, but I thought it might be useful to take in the field for quick measurements on portable ops. But I don't trust it at all yet. Need to figure out what I am doing wrong.

--
Paul W1ip

I saw some on thingiverse but decided that none of them had any strain relief for the connectors. From my experience it is the parts that move that cause the most problems.. I designed a two piece case that clamps the sma connectors. I put in slot to act as a strain relief in it to run the cables to the other side.

I use both 3d design and the nanovna as part of my hobby skills.

Would be glad to print a case for 10 dollars. I also print keys.



73 Dave

Herb,
There are many ways to install Windows drivers. The procedure I wrote works fine for Windows 7 and is repeatable. Unfortunately there are also many variables in all this and each computer environment is different.?
Zadig is just an alternative to install the drivers but is absolutely not a requirement. In fact, it can screw up the drivers if you're not careful using it.?
As the forum member stated, the STM drivers were installed using Windows update.?

... Larry

On Sat, Nov 30, 2019 at 11:52 PM, Roger Need wrote:

… I got it working. I am leaving the answer here in case
=============================================================

So installing the STM32 Virtual COM Port Driver from as suggested by this group's Wiki page did not work for you? If so then the Wiki needs to be updated by adding that Zadig is also required to properly install the serial driver for Windows versions below 10.

- Herb

Thanks everybody. I got it working. I am leaving the answer here in case someone else with Windows 8.1 has the same problem.

When I plug in the USB you hear the Windows sound and see the driver trying to install but it fails. This can be seen in Device manager as an unknown device.

There are several ways to get the driver installed. One is by downloading Zadig. You run it and it finds the unknown device automatically. Then you select the Serial driver from the drop down list of driver options. Click install and now a virtual Com port driver is found under Ports.

The other way works very well too. Just run Windows Update manually and it finds the ST driver as an "Important Update". Click install and in a few seconds a virtual COM pot is ready for use. I think this is the same driver that is supplied built-in to the Windows 10 distribution.

To check that all is OK I ran NanoVNA Saver program and it found the device and I was able to connect and communicate with the NanoVNA.

On 11/30/19 1:02 AM, Gyula Molnar wrote:

do you think everyone is a pc guru?

???? I don't think everybody here is a PC guru, and because they aren't, they should be using Linux so they wouldn't have to worry about drivers.? Or rebuilding their computers after an update, or any of that other Windows stuff.

The only one that makes a greter than 3Ah (3350 mah) 18650 is panasonic Sanyo.

However there are real and good 2200, 2600 and 2800 cells I use the LG2800mah
cells and they are real. However I've seen 2packs of 18650s claiming 4400 mah,
and they were in parallel the cells alone tested at 2200.

With vapers hacking their thing there are a lot of junk cells with outrageous claims.

As to cell size I have tried 18650 at 2800 mah(LG) and it charges properly and
takes about 3 hours. I also tried a Sony 4300mah pouch and it was happy
though it took 5 hours to charge and bulky. Something n the 2100 mah range
with a reasonable form factor is good. However I'm using a set of 1100 mah
cells from a old camera so they can be charged separate if I care to. They
give good run time, charges fast and no noticable battery heating.

It doesn't seem very critical..

-----------------
I do not accept private email due to forum scraping groups.io

This is all fine for accessing the mag of s21. However, I also need its phase. I have not gone through the arithmetic but my concern is diluting the phase term. Dynamic range and a degree of uncertainty may render angles in jeopardy.

Alan

Check you local library - some have 3d printers for public use - at mine, just take in a thumb drive with the files or give them the URL and they will download and they will print it out for free. Neet way to make an iambic key too.

Just a note to say thanks to this group.

It's EXTREMELY rare that a handful of pure enthusiasts set up a group, even rarer that the same bunch within a few months
can get together some GOOD Wikipages and files areas.

I've been using mailing lists and groups for at least 30 years.

Landline BBS, packet radio, Mailman list, Yahoo and now groups.io so I think I've seen a lot of historical usage.

(I've seen some hysterical usage too) ;-))

This group lands itself in the 1% category of excellence, when compared to anything else I have seen, all due
to a handful of *enthusiasts* and within only a few months.

To use the English vernacular, compared to rest, it the dogs bollocks.

- Andy -