开云体育

In your #3, what you're seeing should be due to the fact that the antenna-VNA assembly is "floating" and is
highly subject to things that are in the near field. Basically, anything that is not virtually invisible
to RF (like teflon, etc) becomes part of the antenna system when it is in the near field and when (for
example) your hand comes very close to, or actually in contact with any part of the antenna-VNA assembly,
you become part of the antenna system. Any "shielding" that is only connected to the shield side of the
antenna or the "ground" trace on the analyzer's PCB just becomes part of the antenna system as well. That
alters the impedance seen by the VNA, and that is reflected in what you see on the display. This is why
designing/evaluating antennas for handheld transceivers is such an "interesting" process!

This reminds me of something I read quite some time ago that the shield side is effectively the "other half of the dipole". Usually the shield of an antenna system would be earthed, but in this case when using the NanoVNA, it would indeed be floating and subject to the influence of outside factors such as the hand effect. I expected the effect when my hand was near the antenna, but puzzled when I got it by touching the NanoVNA. This is beggining to make sense. Oh, and I will make sure not to do any antenna testing in the kitchen ! :-) Lots of Teflon and metal in there....

With regards to the 'bdist_wheel' problem, mihtjel has replied very quickly indeed! He advises that on Mint, another dependency is required as it is not installed by default with python. An additional package needs to be installed:

python3-wheel

I installed it as he suggested and the pip install completed without error and the application even appears to start faster. I am grateful to him for looking into the matter and making this application available to NanoVNA users. It is so much easier to drive the NanoVNA from a bigger screen!

I will now wait for all my additonal bits to arrive including the SMA 50 ohm load and the stylus. The SMA 50 ohm load will remove the need for any adapters and cables and I can then run through the calibration steps again to see if that improves things. With the stylus I should be able to move the markers without my hand having to be quite as close to the unit. Not sure what to do about the switch yet, but maybe with the stylus it won't matter.

Thank you for all the responses.

On Sun, Oct 27, 2019 at 11:40 AM, <subs@...> wrote:

that the shield side is effectively the "other half of the dipole". Usually
the shield of an antenna system would be earthed, but in this case when using
the NanoVNA, it would indeed be floating and subject to the influence of
outside factors such as the hand effect.

This is an important 'Gotcha' that catches out many people when measuring antennas.

They disconnect the feeder and measure the antenna at the feedpoint, then wonder why they get unexpected results.

If you are measuring a coax fed antenna and using the VNA at the feedpoint, make sure you connect the outer of the coax to the outer of the VNA SMA connector so that it's more representative of the actual antenna system.

If you are using a balanced antenna with a balun that is working effectively, you should not see any difference between sweeps with the coax screen connected or disconnected.

The Nano VNA is small enough and sufficiently self contained to appear as a balanced load at low frequencies. However as soon as you connect it to a PC, which may in turn be powered from a grounded mains supply, you have an unbalanced load.

You always need to think, "is the antenna I'm measuring configured as it would be when in normal use or am I adding (or leaving out) something in the system that makes it different".

Regards,

Martin - G8JNJ

Well this is embarrassing! It turns out that an SMA 50ohm load and through connector were included in the package that I purchased. They were attached to one of the short cables and I didn't notice until it fell of just now! All the bits I have just ordered will still come in handy anyway so its not been a wasted purchase. I have just gone through the all calibration steps again as described and using the SMA parts supplied. It does seem to work better now and is less prone to the hand effect.

The Nano VNA is small enough and sufficiently self contained to appear as a balanced load at low
frequencies. However as soon as you connect it to a PC, which may in turn be powered from a grounded mains
supply, you have an unbalanced load.

You always need to think, "is the antenna I'm measuring configured as it would be when in normal use or am > I adding (or leaving out) something in the system that makes it different".

Martin G8JNJ, thanks for these points, which do sound important to keep in mind. The SMA connector shield is not isolated from the USB ground so connecting the nanoVNA to the PC would indeed ground the shield side which means the nanoVNA would no longer be floating. That is going to affect the measurement of a balanced antenna system and perhaps even an an unbalanced one.

If you are measuring a coax fed antenna and using the VNA at the feedpoint, make sure you connect the outer
of the coax to the outer of the VNA SMA connector so that it's more representative of the actual antenna
system.

Its a fair point to keep in mind, but in the case of a handheld scanner antenna, wouldn't that always be the case, even when using adapters, unless the connector/adapter were faulty in some way (short/open)? Since the antenna usually sits directly on the BNC connector of the radio when in use, wouldn't measuring at the feedpoint (i.e. the antenna BNC connector) in this case be representative? Naturally if I were measuring a co-ax fed base or car mounted antenna, then I would need to take into account the whole system including feeder cable and make sure that the connects were the correct way around as you point out.

When measuring earlier, I added a 20cm length of thin SMA to BNC co-ax and a BNC t-piece so I guess that would not have been entirely representative? I wonder whether adding those components also skewed the calibration somewhat as it does seem better now that I have used the supplied SMB parts.

The SMD 50 ohm loads arrived today. They measure 51.15 ohms each (I screwed through connectors in to make sure the cap was making firm contact with the barrel) which is about 2.3% out. I guess they used 5% tolerance resistors which for the price is perhaps not surprising. I was hoping to get something within 1% though. The one that was supplied with the NanoVNA measures 49.995 ohm. Measurements take into account the resistance of the DMM probes.

NanoVNA Web client

After most recent Chromebook updates,
NanoVNA Web client now works with Acer CB5-571 Chromebook 15.

Thanks for the tip! Unfortunately the off-line APK version still does not recognize the NanoVNA. At least the on-line version allows Chromebooks with the latest OS to attach to the NanoVNA.

Samsung XE500C13 Chromebook 13

- Herb

Unfortunately the off-line APK version still does not recognize the NanoVNA.

Experimenting with other Android apps using USB OTG,
the problem appears to be ChromeBook Android emulation,
perhaps simply USB dev path naming or permissions.

No blame to cho45

Herb,
I have no issues with the latest APK on my LG tablet (Android 7) or Chinese cellphone (Android 8.1).
Install the following USB utility and check to see if your Android device detects the nano:


Plug in your Nano and check the peripheral section. It should show everything connected.
Regards,??
Larry

Larry,
Thanks for the suggestion. AID64 on the my Chromebook also shows "No USB devices found." with the NanoVNA connected. Oristo thinks android emulation for some USB devices is a work in progress on current Chromebooks. The WebApp works so the hardware interface between the Chromebook and NanoVNA is functional.

- Herb