Re: Some test results using a NanoVNA
One addendum regarding inductance.? I did mention that 50 kHz as a low limit prevents measurement of large inductances.
I just measured a few and got good results up to around 50 mH.? My inductance stock doesn't have specially made parts so the higher inductances have very low self resonat frequencies.? Measuring the bigger ones gives an inductance reading that is too small, since the distributed capacitance begins to shunt the inductance.
So for L up to a few dozen millihenries it's usable but not terribly accurate.? Down to the microhenries and below it's good.? Inductance of a straight wire is easily measured if the test fixture can accommodate it.? Shorting the leads gives a residual reading you can subtract but it's not that simple, as we are dealing with magnetic fields and any change in position messes it up.
The same is true for capacitances below several dozen picofarads.
Bob
toggle quoted message
Show quoted text
On Tuesday, December 17, 2019, 10:12:15 AM PST, Bob Albert via Groups.Io <bob91343@...> wrote: How to measure inductance?? Very simple.? First, you make up an adapter so you can connect the unknown to the nano S0. Calibrate the nano.? The next step depends on the inductance.? Select a frequency appropriate to the part, 50 kHz for low frequency parts and up into the high frequencies for rf chokes and into the VHF and UHF for straight pieces of wire, etc.? Since you can't go below 50 kHz that limits how large an inductor you can measure. Select Smith chart for display.? Read inductance off the screen.? You can read the resistance and compute the Q.? If the inductance is very small, short the leads and take a reading, which you subtract from the final reading. With the leads shorted, the Smith chart should have a dot at the left edge.? With the leads open, at the right edge.? If the dot is not at the edge it indicates a low Q. As a reality check, open up the span to a reasonable frequency range and move the marker over the range.? The inductance should remain relatively constant.? If you open up the frequency range the chart shows a circle.? At some point the circle intersects the horizontal axis at the self resonant frequency.? With high Q you need to keep the span narrow, as the resonanct frequency can elude measurement since it's very sharp. The same procedure works for other components.? For cables, it will show electrical length.? The way you measure that is to adjust the frequency so that the Smith circle, for an open end coax, starts at the right and sweeps down and around up to the left.? Where it reaches the center line on the left is the frequency where the coax is one quarter wave long.? The deviation from the outside circle indicates loss. Then you put a small pot as a termination for the open coax and adjust it until the Smith trace shrinks to a dot in the center.? At that point, measure the pot with an Ohmmeter and it will be the characteristic impedance. This inexpensive device is not a toy.? It's a sophisticated, well designed piece of first rate test equipment.? No ham with tools should be without one. Each time I use mine I learn new ways to use it.? You can measure diode capacitance and, with care, can plot a curve of capacitance vs bias to characterize it as a voltage variable capacitor to use in a PLL or tuning network.? Just don't apply dc to the vna port.? Or any voltage.? It has its own generator, very accurate.? You can measure transistor and tube capacitances as well.? Or connectors. If you connect nothing to the nano it will show a residual reading, sometimes into the femtofarads (one femtofarad is a thousandth of a picofarad).? I never measured a femtofarad before. I measure crystals this way.? This is tricky, as crystals have very high Q and you need to keep the span very tiny.? You need to know the frequency, which is generally marked on the crystal.? Set the nano as that number for center and a small span, say a few kHz.? The crystal frequency will show at the left edge but not on a circle due to the lack of frequency resolution.? Keep narrowing the span until you can read frequency as close as you like, although the resolution isn't all that small. When measuring capacitors the ESR is indicated on the screen.? To measure ESL you measure the self resonant frequency (curve intersecting horizontal line) and compute it from the classic formula.? Or go to a somewhat higher frequency and read the screen.? Remember that the reading is based on a simulation of the part as a resistance in series with a reactance. You measure the distributed capacitance and ESL of a resistor the same way.? With care you can get readings that are very accurate.? The standard used for measuring all of this is the 50 Ohm load you use during calibration so that needs to be very good for precise readings.? My resistor readings are within about 1% of what my GR bridge and HP 3456A show. I know the question was how to measure inductance but I got carried away. Questions? Bob K6DDX ? ? On Tuesday, December 17, 2019, 04:59:21 AM PST, Oristo <ormpoa@...> wrote: > What's needed to connect my NanoVNA- to my Android phone to use the NanoVNA- Webb app? Just a usbc to usbc cable? /g/nanovna-users/message/4970
|
I have read numerous comments that state that the quality of the cables supplied with the NanoVNA is poor. However Hugen has supplied good calibration terminators in his NanoVNA-H in the gift box package. Has anyone tested the cables in the NanoVNA-H gift box to determine their quality and impedance?
Any recommended cables from suppliers in North America?
|
Re: errors of "error" models
|
Re: First PCB pictures of the V2
Great stuff. Sorry if this has been answered elsewhere but are there plans for Bluetooth?
Thanks,
Brian WB8AM
|
Re: First PCB pictures of the V2
Thank you Herb,
So much traffic, I missed the specs.
Dale W4OP
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of hwalker Sent: Tuesday, December 17, 2019 1:29 PM To: [email protected]Subject: Re: [nanovna-users] First PCB pictures of the V2 On Tue, Dec 17, 2019 at 10:10 AM, Dale Parfitt wrote: How does this differ from the 4.3" NanoVNA-F already available? ==================================================================== Recapping from information info previously provided in this thread and elsewhere: 1. Frequency range expected to be 50kHz to 3GHz. 2. Typical S11 floor is -50dB up to 2GHz and -40dB up to 3GHz 3. -70dB S21 floor below 2GHz and -60dB floor up to 3GHz 4. 4 inch display currently being prototyped - Herb
|
Re: First PCB pictures of the V2
On Tue, Dec 17, 2019 at 10:10 AM, Dale Parfitt wrote:
How does this differ from the 4.3" NanoVNA-F already available?
====================================================================
Recapping from information info previously provided in this thread and elsewhere:
1. Frequency range expected to be 50kHz to 3GHz.
2. Typical S11 floor is -50dB up to 2GHz and -40dB up to 3GHz
3. -70dB S21 floor below 2GHz and -60dB floor up to 3GHz
4. 4 inch display currently being prototyped
- Herb
|
Re: Are any of the NanoVNA sold on Amazon any better or worse than any others? Is there a better U.S. site to order from?
On Tue, Dec 17, 2019 at 05:20 AM, Dr. David Kirkby, Kirkby Microwave Ltd wrote:
With a well written specification, following closely the format of a major VNA manufacturer like Keysight, some users would be in a position to determine if the specifications are poorer. Then poor sellers would soon get a bad reputation.
===============================================================================================
In my opinion the vast majority of hobbyists that the NanoVNA-H are targeted to are primarily concerned with frequency range, return loss, and thru loss specifications. Manufacturers of these hobbyist devices have limited resources. Detailed specifications as produced by a major manufacturer such as Keysight are probably best left to the community of open source users who in a lot of cases have better equipment and know-how to produce these specifications.
The NanoVNA-H has a track record of publicly releasing its schematics and frequently releasing firmware updates in response to user feedback. The NanoVNA-F, which has been out for a while now, has chosen not to publicly release its schematics, has not had any significant firmware upgrades since its release, and has only a quick start guide for documentation. Yet the owners of the NanoVNA-F seem willing to overlook these drawbacks because of larger display size and other considerations, such as metal case and easier firmware upgrade. That is a good example of what specifications are really important to the majority of buyers of these low cost VNA's.
- Herb
|
Re: Some test results using a NanoVNA
How to measure inductance?? Very simple.? First, you make up an adapter so you can connect the unknown to the nano S0. Calibrate the nano.? The next step depends on the inductance.? Select a frequency appropriate to the part, 50 kHz for low frequency parts and up into the high frequencies for rf chokes and into the VHF and UHF for straight pieces of wire, etc.? Since you can't go below 50 kHz that limits how large an inductor you can measure. Select Smith chart for display.? Read inductance off the screen.? You can read the resistance and compute the Q.? If the inductance is very small, short the leads and take a reading, which you subtract from the final reading. With the leads shorted, the Smith chart should have a dot at the left edge.? With the leads open, at the right edge.? If the dot is not at the edge it indicates a low Q. As a reality check, open up the span to a reasonable frequency range and move the marker over the range.? The inductance should remain relatively constant.? If you open up the frequency range the chart shows a circle.? At some point the circle intersects the horizontal axis at the self resonant frequency.? With high Q you need to keep the span narrow, as the resonanct frequency can elude measurement since it's very sharp. The same procedure works for other components.? For cables, it will show electrical length.? The way you measure that is to adjust the frequency so that the Smith circle, for an open end coax, starts at the right and sweeps down and around up to the left.? Where it reaches the center line on the left is the frequency where the coax is one quarter wave long.? The deviation from the outside circle indicates loss. Then you put a small pot as a termination for the open coax and adjust it until the Smith trace shrinks to a dot in the center.? At that point, measure the pot with an Ohmmeter and it will be the characteristic impedance. This inexpensive device is not a toy.? It's a sophisticated, well designed piece of first rate test equipment.? No ham with tools should be without one. Each time I use mine I learn new ways to use it.? You can measure diode capacitance and, with care, can plot a curve of capacitance vs bias to characterize it as a voltage variable capacitor to use in a PLL or tuning network.? Just don't apply dc to the vna port.? Or any voltage.? It has its own generator, very accurate.? You can measure transistor and tube capacitances as well.? Or connectors. If you connect nothing to the nano it will show a residual reading, sometimes into the femtofarads (one femtofarad is a thousandth of a picofarad).? I never measured a femtofarad before. I measure crystals this way.? This is tricky, as crystals have very high Q and you need to keep the span very tiny.? You need to know the frequency, which is generally marked on the crystal.? Set the nano as that number for center and a small span, say a few kHz.? The crystal frequency will show at the left edge but not on a circle due to the lack of frequency resolution.? Keep narrowing the span until you can read frequency as close as you like, although the resolution isn't all that small. When measuring capacitors the ESR is indicated on the screen.? To measure ESL you measure the self resonant frequency (curve intersecting horizontal line) and compute it from the classic formula.? Or go to a somewhat higher frequency and read the screen.? Remember that the reading is based on a simulation of the part as a resistance in series with a reactance. You measure the distributed capacitance and ESL of a resistor the same way.? With care you can get readings that are very accurate.? The standard used for measuring all of this is the 50 Ohm load you use during calibration so that needs to be very good for precise readings.? My resistor readings are within about 1% of what my GR bridge and HP 3456A show. I know the question was how to measure inductance but I got carried away. Questions? Bob K6DDX On Tuesday, December 17, 2019, 04:59:21 AM PST, Oristo <ormpoa@...> wrote: > What's needed to connect my NanoVNA- to my Android phone to use the NanoVNA- Webb app? Just a usbc to usbc cable? /g/nanovna-users/message/4970
|
Re: First PCB pictures of the V2
How does this differ from the 4.3" NanoVNA-F already available?
Tnx,
Dale W4OP
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of John Ackermann N8UR Sent: Tuesday, December 17, 2019 1:05 PM To: [email protected]Subject: Re: [nanovna-users] First PCB pictures of the V2 Just another voice asking for a large screen version -- I would really like that! Thanks for all you're doing! John ---- On 12/17/19 11:44 AM, Gabriel Tenma White wrote: shields: I'll be characterizing the sensitivity to external interference before making a final decision on which shields to include.
jog switch: in our tests the switch's position didn't affect usability with cables connected. The switch is less likely to be damaged because it does not protrude as much as it does on the original Nano, and in this position the RF connectors actually protect it from being jammed.
firmware upgrade: the F303 used in the V2 does not have a built in bootloader, so we will look into choices of bootloaders. Most likely the upgrade process will be similar to the -F, but I can't make any promises yet.
display: The ST7796S is command-compatible with the ILI9341 so menu-driven switchover may be doable. I can see an image when I swap a ST7796S display in without any code changes, just in a corner of the screen and flipped. The team will make a decision about whether to offer two versions for sale.
|
Re: First PCB pictures of the V2
Just another voice asking for a large screen version -- I would really
like that!
Thanks for all you're doing!
John
----
toggle quoted message
Show quoted text
On 12/17/19 11:44 AM, Gabriel Tenma White wrote: shields: I'll be characterizing the sensitivity to external interference before making a final decision on which shields to include.
jog switch: in our tests the switch's position didn't affect usability with cables connected. The switch is less likely to be damaged because it does not protrude as much as it does on the original Nano, and in this position the RF connectors actually protect it from being jammed.
firmware upgrade: the F303 used in the V2 does not have a built in bootloader, so we will look into choices of bootloaders. Most likely the upgrade process will be similar to the -F, but I can't make any promises yet.
display: The ST7796S is command-compatible with the ILI9341 so menu-driven switchover may be doable. I can see an image when I swap a ST7796S display in without any code changes, just in a corner of the screen and flipped. The team will make a decision about whether to offer two versions for sale.
|
Re: First PCB pictures of the V2
Good job, perhaps offer the option to rotate display 180 degrees so connectors and jog switch point away from user.
|
Re: First PCB pictures of the V2
Nice job! I can't wait to get one, I could really use one for BLE antenna development. Can I be a beta tester?
Op 17-12-2019 om 15:32 schreef Gabriel Tenma White:
toggle quoted message
Show quoted text
Here are the final pre-production prototypes, and plots showing uncalibrated and calibrated S21 floor. PCB files will be released when V2 is available for order. The LCD connector can accept the 4 inch ST7796S LCD as well, and firmware support for that will be done before product release.
Note that not all of the shield can positions will be populated in production: some shields positions are found to give no performance improvement. The characterization of S21 floor was done with only 4 shields populated. If you get a unit with only 2 to 4 shields populated it's not necessarily a fake!
|
Re: First PCB pictures of the V2
I wish you had chosen an alternate location for the selection switch. Right
between the connectors will be awkward to use. It probably would have been
better to swap the switch and USB connector positions.
Sam Reaves
ARS W3OHM
|
Re: First PCB pictures of the V2
Looks good. What is the lowest frequency you can set the unit to?
Sam
W3OHM
|
Re: First PCB pictures of the V2
Ouch, the toggle switch between the outputs? Inconvenient to say at least. I would hope you find a better switch than this one it is to easy damaged especially in the position it is in. Will you be offering 2 versions or will you let the user decide to upgrade to the larger screen? How will you handle the switch over? Menu driven or a software change? Also, are you keeping the convoluted way the regular nano is using for software upgrade or do you use the way the -F is doing it? If it is done the old way the unit has only limited value to most people as the majority of the people will never upgrade the unit.
-Bilbo
|
Re: errors of "error" models
Good morning GIN&PEZ;
To the extent that I can contribute in some positive manner, I am certainly willing to do so under the careful scrutiny of your editorial council to prevent propagating incorrectly interpreted understanding. I have already been sufficiently rewarded by your patience and willingness to reach out to the amateur radio community, and myself in particular, as an independent resource to challenge the results of your research. It is an honor to be invited as a contributor, and I will endeavor to become credit worthy.
Please keep in mind that my knowledge of your work is continuing to evolve; and my resources here are limited. Exhaustive and testing with precise accuracy is likely out of my reach. You have been warned. :-)
--
73
Gary, N3GO
|
Re: Are any of the NanoVNA sold on Amazon any better or worse than any others? Is there a better U.S. site to order from?
I would strongly advise purchasing for Hugen's site:
<>
I haven't had any issues ordering form the alababa site and the delivery to me in the U.S. was prompt.
Purchasing from Hugen will help better characterize the performance of the nanoVNA if we have the same model.
The nanoVNA-H, and nanoVNA-F, have an excellent price-to-performance ratio for a piece of hobbyist test equipment. We should support the original developers.
Mike N2MS
|
Re: First PCB pictures of the V2
From: Gabriel Tenma White
Subject: [nanovna-users] First PCB pictures of the V2
Here are the final pre-production prototypes, and plots showing uncalibrated and calibrated S21 floor. PCB files will be released when V2 is available for order. The LCD connector can accept the 4 inch ST7796S LCD as well, and firmware support for that will be done before product release.
Note that not all of the shield can positions will be populated in production: some shields positions are found to give no performance improvement. The characterization of S21 floor was done with only 4 shields populated. If you get a unit with only 2 to 4 shields populated it's not necessarily a fake!
============================================
Gabriel,
This looks most encouraging - especially if the 3 GHz performance it as shown! I would certainly be tempted to get one (it would be my third) but only if it comes in a boxed unit, with the same 4.3-inch display or similar. I.e. a 3 GHz version of the NanoVNA-F.
Congratulations!
David
--
SatSignal Software - Quality software for you
Web:
Email: david-taylor@...
Twitter: @gm8arv
|
Re: First PCB pictures of the V2
On Tue, 17 Dec 2019 06:32:32 -0800, you wrote: Here are the final pre-production prototypes, and plots showing uncalibrated and calibrated S21 floor. PCB files will be released when V2 is available for order. The LCD connector can accept the 4 inch ST7796S LCD as well, and firmware support for that will be done before product release.
Note that not all of the shield can positions will be populated in production: some shields positions are found to give no performance improvement. The characterization of S21 floor was done with only 4 shields populated. If you get a unit with only 2 to 4 shields populated it's not necessarily a fake!
Shields will only show performance in a RF environment, if your lab is clean, you will not notice anything. So keep the shields in place, please. Going to become a VolksVNA. OE8UWW
|
Re: Are any of the NanoVNA sold on Amazon any better or worse than any others? Is there a better U.S. site to order from?
On Tue, 17 Dec 2019 09:25:31 -0500, you wrote: One thing I do not understand is why connecting the port 1 and 2 grounds
together should decrease performance. Shouldn't all grounds be connected
together, and all of these grounds to the chassis? Every piece of RF
equipment I have seen that has a metal chassis grounds the RF to the
chassis in as many places as possible. The chassis both serves as a
constant potential large capacitive body and as a RF shield. The nanovna
PCBs I have seen for some reason have separation of the ground planes
between port 1 and 2. I agree there should be a via fence between the
ports, but shouldn't the ground be common between the two with no break in
the ground plane? This is a feature of the edy555 design I do not
understand. Leaving these two with a large break in the ground plane is
basically having port 1 and port 2 ground pads acting as two sides of an
antenna to pick up stray fields.
I was wondering about that also. There are two ways to go: (simplified) 1.) Star-grounding usually with audio to avoid hum and with low voltage thermocouples to avoid ground loops. 2.) Brute-force ground-it-all, usually by applying ground layers on both sides of the print and through-hole riveting every millimeter. Best for HF stuff. The two inputs are ground connected already over the print ground layer. OE8UWW
|
Roger Need
gin&pez@arg
Brian Garber
Sam Reaves
David J Taylor