开云体育

Roland,
Are there any circulators included in the duplexfilter?
--John Gord

Erik,
Even the expensive measurement equipment that I've used in my occupation had mismatch problems, and the advice was always to use attenuators on the transmit and receive ports to reduce the mismatch errors. I guess it depends on what your application for the NanoVNA is and how much of a mismatch error your application can tolerate. In the 50k - 300 MHz range you can use a 10 dB attenuator on each NanoVNA port and still have at least 50 dB of sensitivity for S21 measurements. Above 300 MHz with 20 dB in-line some S21 measurements will lose usefulness.

Expensive lab equipment can easily take 10 dB on each port and still give 80 dB or more of dynamic range up to 1 GHz. For $50 I don't sweat the NanoVNA mismatch errors, although out of force of habit, I do use 3 dB 18 GHz HP attenuators on each port when I have a lot of headroom to work with.

- Herb

Hi
This is a 3 port device. I assume the unterminated port during your alignment is terminated into 50 ohms.

Alan

Hi Erik
I seems you have made some error correction one way or the other. The S11 ImagZ in the first image attached is inductive as it should be and nicely straight up to 300MHz. You also see the S11 smith chart trace follow the SWR 2 circle nicely. In the second image the S21 trace changed to continuous phase and I have implemented a port extension of 141ps, which is the delay introduced by the T-adaptor and you see the T-Check does not change (and it shall not do), and the section of the trace in the Smith Chart running backward is the points from 300MHz upwards not having a delay of 141ps, due to some imaginary elements in the path Ch0 to Ch1 not directly to identify but could also be some internal errors in the NanoVNA which a proper error correction should fix . The ImagZ is now dropping with the port extension implemented indication it is a capacitiv shunt of some sort which might be the shunt C in your loads as it start from 50KHz. It can also be the open cal standard not correctly set and in the nanoVNA-saver you can see the effect of changing these parameter live during continuous sweeping.
That is the analyzes I can extract.
How did you obtain this fine result ? introducing error corrections ??
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af erik@...
Sendt: 9. november 2019 19:47
Til: [email protected]
Emne: Re: [nanovna-users] T-Check for my nanoVNA - Results look excellent below 150 MHz and acceptable up to 300 MHz

Here my latest T-check measurement with nanoVNA firmware adapted for better measurement performance and a 10dB attenuator directly after the Tee (calibrated out in the thru calibration) The small jump at 300MHz is when the nanoVNA switches to harmonics mode.
The drift upwards above 600MHz is probably due to the previously mentioned non-linearity so possibly impossible to improve.
If you look really careful with a lot of imagination you can see that the pattern below 300MHz is amplified visible between 300 and 900MHz. My current theory is that non-linearities are amplified when goin to harmonics mode but already present in the fundamental model. I see similar behavior in my home build nanoVNA equivalent (also uses SI5351 and SA612) but nothing like this in my home build 3Ghz VNA Maybe some more experiments with further reducing the driving power, while also solving the noise problem, may provide some evidence

@Kurt, I would not mind if you have a look at the s2p file to check if I did not make calculation mistakes....
--
Erik, PD0EK

Hello,

I've got a question.
I have a Celwave duplexfilter for UHF. I tuned it for 7,6MHz spacing.
Tuning is done with the nanoVNA and patience.

But... I discovered something strange.
The antennaport is connected to CH1.
The rx or tx ports I connected to CH0.

If I try it backwords, antenna on CH0 and tx/rx on CH1, I get completely other measurments.
It is with this settings impossible (?) to adjust the filter.

Can anybody tell me what the right setting is for tuning a duplexfilter?

Thanks in advance,
Roland.

Is a working link.

I just found an Android USB terminal app that works with the nanoVNA on my tablet.?
?
I've attached a screenshot. You need to use the CDC driver.?






On Sat, 9 Nov 2019 at 2:00 PM, Oristo<ormpoa@...> wrote: > Android 5 is too old - I believe you need 7 and up for this.

The Android app has JavaScript ported from the web app,
using Capacitor?
.. which requires at least Android version 7.0 on API 24 to use the Android Emulator.

That web page goes on to suggest that "Physical devices should work
as low as Android 5.0 (API 21) as long as their System WebView is updated."
User comments for the update app are discouraging:


.. and anyway my newest Android device is only 4.2.2 (API 17)
and despite its USB OTG working with some other USB devices,
all six serial terminal Android apps tested hung the nanoVNA.

Hi Rudi

Maybe one of these days I will proof it is hopeless ?

Kind regards

Kurt



-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af reuterr@...
Sendt: 9. november 2019 19:41
Til: [email protected]
Emne: Re: [nanovna-users] T-Check for my nanoVNA - Results look excellent below 150 MHz and acceptable up to 300 MHz



Hello Kurt,



Thank you very much for your recommendations, but I did not make any progress today, so I give up.



I think the problem is in the CH1 bridge impedance, and I do no longer want to try to improve the T-Check result.



It is not my target to make the NanoVNA comparable to a HP VNA.

I have seen my limits, thats enough.

My curiosity is satisfied, maybe someone more experienced finds a way to improve with the T-Check the NanoVNA.



How about you Kurt ;-)



73, Rudi DL5FA

Android 5 is too old - I believe you need 7 and up for this.

The Android app has JavaScript ported from the web app,
using Capacitor
.. which requires at least Android version 7.0 on API 24 to use the Android Emulator.

That web page goes on to suggest that "Physical devices should work
as low as Android 5.0 (API 21) as long as their System WebView is updated."
User comments for the update app are discouraging:


.. and anyway my newest Android device is only 4.2.2 (API 17)
and despite its USB OTG working with some other USB devices,
all six serial terminal Android apps tested hung the nanoVNA.

Hello Erik,

One problem with the nanoVNA is the far from perfect match of port 2.

I have seen your post too late, to include it in my post for Kurt.

It makes me happy to see, that you came to the same conclusion like me :-)

73, Rudi DL5FA

Here my latest T-check measurement with nanoVNA firmware adapted for better measurement performance and a 10dB attenuator directly after the Tee (calibrated out in the thru calibration)
The small jump at 300MHz is when the nanoVNA switches to harmonics mode.
The drift upwards above 600MHz is probably due to the previously mentioned non-linearity so possibly impossible to improve.
If you look really careful with a lot of imagination you can see that the pattern below 300MHz is amplified visible between 300 and 900MHz. My current theory is that non-linearities are amplified when goin to harmonics mode but already present in the fundamental model. I see similar behavior in my home build nanoVNA equivalent (also uses SI5351 and SA612) but nothing like this in my home build 3Ghz VNA
Maybe some more experiments with further reducing the driving power, while also solving the noise problem, may provide some evidence

@Kurt, I would not mind if you have a look at the s2p file to check if I did not make calculation mistakes....
--
Erik, PD0EK

Hello Kurt,

Thank you very much for your recommendations,
but I did not make any progress today, so I give up.

I think the problem is in the CH1 bridge impedance,
and I do no longer want to try to improve the T-Check result.

It is not my target to make the NanoVNA comparable to a HP VNA.
I have seen my limits, thats enough.
My curiosity is satisfied, maybe someone more experienced
finds a way to improve with the T-Check the NanoVNA.

How about you Kurt ;-)

73, Rudi DL5FA

Hi Jos,
Grab a copy of the Console Commands document. It now has a section at the end that lists and links to the major firmware developers for the NanoVNA.??
Go to any of the GitHub locations and click on the releases link, then download the latest firmware. Each developer has some features that are unique only to their release but the basic operation is the same across all releases.?
Then, grab a copy of the user guide and read the sections on converting hex files to DFU files (in windows). It's very easy.?
... Larry



On Sat, 9 Nov 2019 at 12:32 PM, Jos Stevens<jrs@...> wrote: Hi Folks,

My name is Jos and I am new here and looking for the best firmware for my NanoVNA device.
I am using my NanoVNA? with the "NanoVNA-Saver" sofware and had problems with downward lines in the returnloss chart.
I reported .thatt at the github/NanoVNASaver site and they said it was firmware related and should use the Version 0.4.0. Firmware.
The only 0,4.0 firmware I could find on the web was on edy555's github site, but the ones I downloaded were of version 2.2? an in .bin.and .hex format..
I can't use these in the? DFUseDemo tool,
So.I went to this group and? looked in "Files/Firmware/ Eriks latest build" area. and the last version there was Version? 0.2.3 in ,fdu format.
I have flashed this firmware in my NanoVNA device now.
After all this, I am a bit confused and my question is :? Does there really exist a version 0.4.0 version by Erik ???
If not why? he calls it version? 0.4.0 on his guthib page ??

Any reaction is appreciated.
Jos

One problem with the nanoVNA is the far from perfect match of port 2.
If you measure the tee with two 50 ohm loads the S11 is a nice line tracking the 25 ohm curve upwards
Once you replace one of the 50 ohm loads with a cable going to port 2 the S11 becomes a wobbly line, just like you see in the T-check calculation.
The attenuator after the tee does help to reduce the mismatching of port 2 but even with 10dB attenuation the mismatch is still visible.
Another problem is the non-linearity that appears above 600MHz. This is possibly caused by leaking of the test signal to the reflection mixer or harmonic generation inside the reflection mixer.
Is this all a problem?
Yes, if you are doing a two port measurement of a device of which the input impedance is highly dependent on the output loading (e.g. port 2) and you are measuring above 500MHz
No, if you are not using port 2 for example if you are doing a pure SWR measurement of a single port device (like an antenna) and you are below 300MHz
--
Erik, PD0EK

The only 0,4.0 firmware I could find on the web was on edy555's github site

Did you download this?

Hi Folks,

My name is Jos and I am new here and looking for the best firmware for my NanoVNA device.
I am using my NanoVNA with the "NanoVNA-Saver" sofware and had problems with downward lines in the returnloss chart.
I reported .thatt at the github/NanoVNASaver site and they said it was firmware related and should use the Version 0.4.0. Firmware.
The only 0,4.0 firmware I could find on the web was on edy555's github site, but the ones I downloaded were of version 2.2 an in .bin.and .hex format..
I can't use these in the DFUseDemo tool,
So.I went to this group and looked in "Files/Firmware/ Eriks latest build" area. and the last version there was Version 0.2.3 in ,fdu format.
I have flashed this firmware in my NanoVNA device now.
After all this, I am a bit confused and my question is : Does there really exist a version 0.4.0 version by Erik ???
If not why he calls it version 0.4.0 on his guthib page ??

Any reaction is appreciated.
Jos

The only little problem I see in reald is it doesn't reverse text colors (fg/bg) to show which trace is selected, as hugen does.

Should be fixed now. Please test.

73 DE DL9CAT

Hi Daniel,
partialy true... ILI 9486/9488 have different key commands compared to 9341. Both does support 16bit pixel data but via parallel interface, via native SPI you can only use a strange 8 bit format or 24bit . Nice SPI interface with 16bit pixel data is available on raspberry pi 3.5/4" modules and have spi to parallel converter on board.
3.5" Raspberry pi lcd is very cheap.

Slawek/SP9BSL

This is fixed by using different firmware

In newer firmware, use shell command `threshold`,
followed by `saveconfig`
/g/nanovna-users/files/NanoVNA_Console_Commands_Nov-4-19.pdf

Datasheet for the si5351 says it works up to 225 MHz.
The original Nano VNA firmware pushed it up to 300 MHz, which has proven to be a wee bit too far.
Above 300 MHz it used the third (and fifth) harmonics of the square wave out of the si5351 instead of the fundamental.
This is fixed by using different firmware, try downloading firmware that claims a maximum of 800 mhz instead of 900 mhz,
This firmware only asks the si5351 to give a max of 800/3 = 267 MHz, which all si5351 chips seem capable of.

Different si5351's crap out at slightly different maximum frequencies.
Just a natural variation due to slight differences between parts,
even if in the same manufacturing batch.

Jerry, KE7ER