开云体育

a 1 meter cable
I have not yet found an error model that can model errors above 300MHz

Kurt suggested that very good cable is wanted, e.g. airline.
I notice better results from LMR-400 than RG58 or RG316

After calibration with the correct calibration error parameters, a 1 meter cable connected to the reference plane with short and open at its other end, should give perfect circles spiraling inwards in the smith chart.
This however was NOT true for my nanoVNA
Using Octave I tried to do a parameter search for the OpenC0,C1 and ShortL0,L1, OpenOffset, ShortOffset, LoadOffset and LoadL0,L1 that would provide the right calibration correction but that did not work. At least the parameter search failed to converge to realistic numbers and the residue was always considerable. Even when only using OpenC0 the residue was large and had way too much structure

So I added a small error model for the bridge with only a constant offset for the reference R, both real and imag)

The first picture shows the open en short errors for the 1 meter cable after calibration without any error correction or calibration model parameters.
The red and green lines are the abs of the deviation of the S11 versus a straight line from 50 to 300MHz The blue line is the abs sum of both errors.
The second picture shows the final modeling parameters and the residual error. the residue is now mostly noise. Only a OpenC0 of 52fF is needed to correct the calibration set to this level. The bridge correction is 0.29 - 0.6i ohm.
The third picture shows the S11 of a 1 meter cable with open end from 0.5 till 900MHz
I have not yet found an error model that can model errors above 300MHz

Any suggestions?

Is there already a place to go for this? I looked at some GitHub pages

QRP RX recently updated his README:

Bob,


But be careful - some vendors charge big $$ for shipping.

Where to buy IP5303?

I didn't have an SMT diode so just used an axial 1N4148.

Thanks for the image, KVR5

Your hand is steadier for soldering than mine..

@QRP - does the small red mark next to it
signify bad hardware?

That's exactly what I do at LC.

"You can also connect trap to hot ends of both CH0 and CH1 and do S21 measure without need for coil adapter. It worked for me when I was unable to detach trap from antenna." - YT9TP, Pedja

IF
your reply DID NOT start with put-down I would reply....

IF you would have looked in the Files and Wiki sections of the forum - all of your questions would have been answered except #1.There is no trademark - only a copyrighted name, owned by edy555.
Check the links to the left of the forum webpage or at the bottom of this reply.
=-=-
Learn about the NanoVNA in the following forum areas:
Documentation & Update Files: /g/nanovna-users/files
Knowledge-based Wiki: /g/nanovna-users/wiki
-=-=-

...Larry

Hmm, putting two signals into a mixer generates more than one single sinusoidal signal.

Sure, but one fixed LPF suffices, rather than LPF tracking swept square wave.

Before purchase I would like to have the following answered / verified
Is “NanoVNA” a trade mark? (verify)
Is “NanoVNA” open source – both hardware and software ? (verify)
“NanoVNA “ is sold by many vendors – is there an official manufacturer?
“NanoVNA “ appears to be “open source “ software - is there an official coordinated effort to keep track of development?
“NanoVNA “ appears to be “open source “ software - is there an official coordinated effort to keep track of “wish list” of new features?

From perpetual / legal “open source” notes in software listing – is this product actually “one man band effort”?

Reasons for asking.
I have been “working” on my own version of VNA , maily because I like to punish myself fighting with hardware and software. I have no commercial interest in this, it is just a hobby.

I am currently stuck developing bluetooth interface to my version of VNA and noticed some interest of this group to have NanoVNA with “wireless capability” .
I could offer some pointers to whomever will eventually (!) add such ability to NanoVNA- internally( hardware) or externally (software) or both . I do not wish post my notes to the group, it generally results in unmanageable e-mail traffic.
Vaclav

On Wed, Oct 30, 2019 at 06:15 PM, Rich NE1EE wrote:

I have only 1 50 O at this point...more on order. I wish that we had a vetted
supplier of low cost terminators.

connect it to CH0, leave CH1 open

On Wed, Oct 30, 2019 at 03:52 PM, <erik@...> wrote:

Unless you want to understand the calibration process do not look at
uncalibrated and uncorrected data

There is a testing of hardware, to make sure that it provides adequate measurements.
With calibration enabled possible hardware issues will be hidden by calibration.
This is why I asked for screenshot with CAL RESET.

On Wed, Oct 30, 2019 at 03:33 PM, Rich NE1EE wrote:

I don't know what to expect, so would appreciate detailed feedback...and if I
didn't do it correctly, please advise, and I'll be happy to redo.

Do the same with load on CH0 and CH1

I can do one port at a time if that is useful.

Cross-talk will change, but better than none.
FWIW, CH1 precision is less critical;
I use uncalibrated 50 Ohm BNC with SMA adapter.

Hi

Hmm, putting two signals into a mixer generates more than one single sinusoidal signal.

Bo

Two adf4351 modules. One at two GHz into LO port of mixer and the other at 2
till 3 GHz through 3 GHz low pass filter an 10 dB attenuator into RF port of
mixer. The IF port should output a acceptable sinus from 0 to 1 GHz

As an alternative to mixing and LPF for lower RF,
grab an eBay AD9959 board:


.. and slave it by SPI to

.. but cost is > $70

On 2019-10-30 08:22:-0700, you wrote:

I don't know why the request was made, and I am pretty new to all this...

Please do, with 50 Ohms directly terminating CH0 and also CH1
The request was for gathering statistics on terminated values before calibration.

I have only 1 50 O at this point...more on order. I wish that we had a vetted supplier of low cost terminators.
I can do one port at a time if that is useful.
Not sure when to expect the others...have not gotten an ETA yet.

Originally part of this thread about S11 open and short Smith chart views

Yes, in the context of being calibrated (with 50 Ohms @ CH0, CH1)

Thanks for that.

73



--
On the banks of the Piscataqua
Rich NE1EE

a LPF a bit above the fundamental frequency can do the job

I think such filters add more cost than starting with sinusoid-generating chips, e.g.
/g/nanovna-users/message/5979

Hi Oristo

A square wave can be expressed as an infinite number of odd harmonic sinusoidal signals. Thus, a LPF a bit above the fundamental frequency can do the job.

Bo