开云体育

Never mind - just click on the "Read-Only" button to open the presentation

Why is it password protected ?

Hello,

This email message is a notification to let you know that the following files have been uploaded to the Files area of the [email protected] group.

• /Smith-Chart Intro_2016_en_pwd.zip

Uploaded By: W1RS <Deflatermaus@...>

Description:
Introduction to the Smith-Chart, (Powerpoint, en)

Cheers,
The Groups.io Team

Here is another good reference
from Fritz Dellsperger, Professor for RF and Microwave Engineering (HB9AJY )
at Bern University of Applied Sciences
from 1992 to 2012, now retired


There is a good powerpoint presentation link in the page. I will try to post it to the Files section as "Smith-Chart Intro_2016_en_pwd.zip"

There is a Computer Smith-Chart Tool and S-Parameter Plot called "Smith" V4.1, The demoversion may be freely distributed.
The demoversion is limited to 5 datapoints and 5 elements which is good for matching design.

On Tue, Mar 31, 2020 at 01:30 PM, Jeff W6FCC wrote:

RG8X

RG8X (foam dielectric) has Vp of 0.79 and loss at 7.2 MHz of 0.75 dB per 100 feet.

Electrical wavelength length of RG8X @ 7.2 MHz is 107.9 feet. One trip around the Smith chart is half wavelength so 100' would be 1.85 rotations around the chart. The round trip loss would be 1.5 dB which is 0.84 reduction in Rho vector that spirals inward as length of coax gets longer and loss increases. So you should push the value point on Smith chart outward (increase number) by 1/0.84 or 1.19 times to correct of loss of coax.

It would be best if you can put a short at the end of the coax to measure the electrical length and loss. If coax is RG8X and 100' in length, and VNA is calibrated without coax, the left most point on Smith chart (minimum Z) should be at 7.77 MHz. At this point the round trip loss should be -1.59 dB ( rho = 0.833) inward from outer ring of Smith chart, at left-most point along the real horizontal center line. (loss at 7.77 MHz is 0.794 dB per 100')

Hi I am new to this group and this topic is kind of old but maybe the problem is still there.
I had the same issue and I got it working in this way: open file setup.py and, at the end, replace

install_requires=[
'pyserial',
'PyQt5',
'numpy',
'scipy'
],

with

install_requires=[
'pyserial',
'PyQt5 <= 5.14.0',
'numpy',
'scipy'
],

It worked for me, I am running Ubuntu 18.04.4 LTS

Paolo - IZ2RXS

Jeff,
Of course the antenna is not 50 Ohms. No antenna is 50 Ohms across the
whole band or bands. Like I said, the coax just rotates the antenna
impedance around the Smith Chart. While you can't measure the impedance
without knowing the exact length and Vf of the transmission line, the
VSWR/Return Loss is the same at both ends of the transmission line minus
twice the insertion loss of the line.
73, Carey, WB4HXE

On Tue, Mar 31, 2020 at 4:27 PM Jeff W6FCC via Groups.Io <w6fcc=
[email protected]> wrote:

Carey,

If the antenna is NOT 50 ohms (it's actually an offset multi-band dipole
with a 4::1 balun transformer), or not resonant, I was thinking that the
COAX would introduce some problems.
They do use small runs of COAX for matching, that's what I was thinking,
that the SWR would not be constant across the length of the COAX if the
antenna was not 50 ohms resonant.

I was looking for a way to put the imaginary connector to the antenna at
the far end of the COAX but maybe that's not going to solve anything.

Jeff

--
Carey Fisher
careyfisher@...

Hi Jeff
If you use the NanoVNA-Saver and do a proper calibration then you just measure at the TX side and save the measurement to a s1p Touchstone file. Then download the ZPlots by AC6LA and open up the s1p file- Then you can shift the measurement the 100 feet to the antenna position.
Following conditions:
The cable type and manufacturer to be selected in ZPlots must be know
The cable length must be known exactly else use the Time domain of NanoVNA-saver but VF must be know for the cable
You need to have Windows EXCEL to run ZPlots
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af Jeff W6FCC via Groups.Io
Sendt: 31. marts 2020 19:31
Til: [email protected]
Emne: [nanovna-users] 100 ft Coax to Dipole Antenna - Can the NanoVNA show just the Dipole analysis?

Setup: I have a 40 meter dipole at 30 feet. Feeding with 100 ft of RG8X
Measurement: 7.200 Mhz, SWR at RADIO end of COAX can be measured (For the Analysis System -- coax + antenna)

Question: Can the Reference plane be "calculated" by software. It's the location of the antenna at the END of the Coax, discounting the Feedline, perhaps by entering the Feedline Impedance and Velocity Factor and Loss?

Asked a different way:

Can the NanoVNA be used to isolate the antenna at the far end of the COAX and somehow IGNORE the impact of the COAX itself ?

I realize the best way to check the antenna is AT the feedpoint, but that is not possible right now. I'm stuck at the Radio End of the COAX and trying to determine of the DIPOLE is too long or short.

One possible "solution" IF I could take a second piece of 100' RG8X and do the calibration at the far end (OPEN, SHORT, LOAD) and push the Reference plane out to the 100' mark.
But can it be done without a second piece of COAX ??

Has anyone calibrated their NanoVNA using a long piece of coax and doing the CAL at the end of the coax before attaching to an antenna to be tested ?

Jeff, W6FCC, 3-31-2020

When I made mine.? I used some jumpers and a bnc-bnc barrel in the middle to make up for the bnc T (about the same length), calibrated with all that attached, then just changed the barrel to the BNC and tuned the stubs.? Worked fine for us for the Idaho QSO party.


KI0E

Bill,

I have not heard it discussed here, but I made a little 'jig' with a series 50 ohm carbon resistor sandwiched between

a male and a female coaxial connector (made one for PL259/SO239 and N-male/bulkhead N-female).

You use your standard 246/f * Vf, but make it a little longer than you think.? Short the far end for even half-wave

multiples, leave open for 1/4 wave, 'TUNE' coax by highest return loss/lowest swr.

This is how I tune the coax for transformers (odd multiple of 1/4 wave) or feed lines for an array (even half wave).

I suppose this could be "fun" trying to do this with sma on the Nano!

73

Glen K4KV

ps - there are other iterations, by doubling or halving the target frequency...

DISlord,

Thank you so much.

Jos

Op 31-3-2020 om 21:23 schreef DiSlord:

Bill, What kind of stubs are you considering? Stubs for Z matching or simple 1/4 wave stubs for filtering or something else?

Yes, you disconnect the antenna from the cable and attach your calibration kit to the free end of the cable with the analyser at the shack end and do a single port calibration. Then reattach the antenna and make your measurements.

Knowing the coax type and length, you can use SimSmith to propagate the
measurement back to the antenna feed point.

I've offered so many emails to the SimSmith download site on this group
that I'll leave it as an exercise to the interested and motivated user to
find the link. It's easy. Just use any search engine.

Dave - W0LEV

On Tue, Mar 31, 2020 at 5:30 PM Jeff W6FCC via Groups.Io <w6fcc=
[email protected]> wrote:

Setup: I have a 40 meter dipole at 30 feet. Feeding with 100 ft of RG8X
Measurement: 7.200 Mhz, SWR at RADIO end of COAX can be measured (For the
Analysis System -- coax + antenna)

Question: Can the Reference plane be "calculated" by software. It's the
location of the antenna at the END of the Coax, discounting the Feedline,
perhaps by entering the Feedline Impedance and Velocity Factor and Loss?

Asked a different way:

Can the NanoVNA be used to isolate the antenna at the far end of the COAX
and somehow IGNORE the impact of the COAX itself ?

I realize the best way to check the antenna is AT the feedpoint, but that
is not possible right now. I'm stuck at the Radio End of the COAX and
trying to determine of the DIPOLE is too long or short.

One possible "solution" IF I could take a second piece of 100' RG8X and do
the calibration at the far end (OPEN, SHORT, LOAD) and push the Reference
plane out to the 100' mark.
But can it be done without a second piece of COAX ??

Has anyone calibrated their NanoVNA using a long piece of coax and doing
the CAL at the end of the coax before attaching to an antenna to be tested ?

Jeff, W6FCC, 3-31-2020

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

Thanks for replying. I thought the vector network analysis would be able to adapt to the cable properties to get an accurate measure of the load at the end of the feedline.

Jeff

I make stubs on a standard bnc T f/m/f connector and use a f/f barrel in
the design, the reason being is that in the field, a wonky system swr will
electrically lengthen the stub and can be compensated by using a bnc f/f/f
gendered T and removing the barrel, assuming you planned ahead and brought
everything.

Or you could always find a trombone player.

The other cabling shouldn't act like a stub: I don't worry about it.

You can only measure the actual antenna impedance remote from the antenna, with a 1/2 wavelenght of coax or feedline (x the velocity factor of the feedline), measured at a single frequency of interest.

If you are going to use a multi-band antenna, you particularly want to measure the impedance at the coax input point in order to know what the transmitter will be actually looking at, as a function of the antenna/feedline system.
Ray,
W4BYG

--
This email has been checked for viruses by AVG.

Carey,

If the antenna is NOT 50 ohms (it's actually an offset multi-band dipole with a 4::1 balun transformer), or not resonant, I was thinking that the COAX would introduce some problems.
They do use small runs of COAX for matching, that's what I was thinking, that the SWR would not be constant across the length of the COAX if the antenna was not 50 ohms resonant.

I was looking for a way to put the imaginary connector to the antenna at the far end of the COAX but maybe that's not going to solve anything.

Jeff

For additional undestand how it work:

You can use generator, set freq for example 20Mhz, set AM modulation 60% and freq 1kHz
Output for example 100mV

Connect to NanoVNA CH1

Set on NanoVNA center freq 20MHz, and span 5kHz, and CH1 Logmag
Now you can see AM signal spectre on you NanoVNA

You can change bandwidth setting and see changes (its like VBW option on any spectrum analyser), less BW freq allow recive less noise and better measure signal data, but need made more measures

DiSlord wrote:
bandwidth N

N - is number from 0 to 255 (not reccomend use values bigger then 99, possibly counter overflow)

N+1 number of samples count passed in dsp
One sample process give 1kHz bandwidth
Two samples give 500Hz
100 samples give 10Hz

BW = 1000 / (N+1) Hz
==============================================

Additionally, for the information of other group members, sending the bandwidth command without any qualifier returns the current bandwidth, i.e.:
sending:
bandwidth
returns
bandwidth 0 (1000Hz)

From the DISPLAY-BANDWIDTH menu only 5 bandwidths are selectible:

1 kHz, 300 Hz, 100 Hz, 30 Hz, 10 Hz

- Herb