开云体育

There has been considerable discussion lately on how to measure resistors, inductors and capacitors using the NanoVNA. In particular measurement errors of these components seems to be a topic of interest. I wrote a technical note on this subject which I hope is useful to members of the group.

You can view and download it from my Box account >>

Roger

The following files and folders have been uploaded to the Files area of the [email protected] group.

* /menu maps/nVNA-H4 v4.3_MS, firmware 1.1, Menu Items & Descriptions 0110..zip ( /g/nanovna-users/files/menu%20maps/nVNA-H4%20v4.3_MS,%20firmware%201.1,%20Menu%20Items%20&%20Descriptions%200110..zip )

*By:* Rich NE1EE <TheDustyKey@...>

*Description:*
Structured text document. Menu descriptions, brief explanations, notes. Spreadsheet used to check some answers. Probably 80% complete. Source docs in MS Word doc and docx. Published doc in PDF.

how is the black dot determined

If there are no black dots, I'd say it's likely a matter of by guess and by gosh. Simple, no?

On Sun, May 8, 2022 at 09:47 AM, Tom VA7TA wrote:

Details of how to make a set of tweezers were published in the Nov/Dec, 2021
issue of ARRL's QEX magazine. Some views of the results from testing SMD
components can be found within the QEX article support files that can be
downloaded from the ARRL QEX files website:

Tom,

Thank you for posting about your NanoVNA tweezers. I downloaded the zip file and looked at your measurements of SMD capacitors and inductors. The plots and tables show some very accurate results for your SMD components across a wide range of frequencies. You also demonstrated that good results can be obtained well away from the +j50 and -j50 ohms of reactance (+/- 90 degree reflection angle) which are the most accurate measurement points. I have also found the NanoVNA to be amazingly accurate for such a low cost device.

Most of us do not have access to QEX so we don't know what your tweezers look like. Could you post a few pictures of your tweezers and test fixture?

Roger

Think you mean *tinySA*, /not/ tinyVNA. (See /g/tinysa and )

Good luck chasing the QRM.

HTH, 73,

Robin, G8DQX

On 5/8/22 11:31 AM, Gregg Messenger VE6WGM wrote:

Thank you for that information Tom.

I have received a fair bit of negative push back regarding my YouTube videos showing the use of alligator clip leads for testing discrete components. My latest video (link shared at the opening of this thread) shows an experiment I did to see what the difference was between a test fixture and alligator leads.

I agree with you regarding maintaining the same physical parameters with the alligator leads as during calibration. My conclusion in a nutshell was that the test fixture was easier to maintain consistency due to the nature of having things solidly mounted in place on a PCB. It is a better way to do it.. however.. clip leads do work very well. My impression was that it basically boils down to what a person wishes to do with the results and what kind of accuracy or repeatability one requires.

I've done ok with clip leads, and taped them to the bench top, so that the spacing remains consistent. It's a sort of "one-off" test fixture.

My 2-cents: Clip leads with proper calibration and attention to
mechanical detail can be used successfully at HF frequencies. From
experience I would put an upper limit on that of 30 MHz. Beyond that, use
a fixture to assure consistent mechanical stability from the cal to
measurement.

Having written that, I usually use a fixture.

Dave - W?LEV

On Sun, May 8, 2022 at 6:31 PM Gregg Messenger VE6WGM <techgreg@...>
wrote:

Thank you for that information Tom.

I have received a fair bit of negative push back regarding my YouTube
videos showing the use of alligator clip leads for testing discrete
components. My latest video (link shared at the opening of this thread)
shows an experiment I did to see what the difference was between a test
fixture and alligator leads.

I agree with you regarding maintaining the same physical parameters with
the alligator leads as during calibration. My conclusion in a nutshell was
that the test fixture was easier to maintain consistency due to the nature
of having things solidly mounted in place on a PCB. It is a better way to
do it.. however.. clip leads do work very well. My impression was that it
basically boils down to what a person wishes to do with the results and
what kind of accuracy or repeatability one requires.


Regards,
Gregg
--
VE6WGM

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


--
Dave - W?LEV

Thank you for that information Tom.

I have received a fair bit of negative push back regarding my YouTube videos showing the use of alligator clip leads for testing discrete components. My latest video (link shared at the opening of this thread) shows an experiment I did to see what the difference was between a test fixture and alligator leads.

I agree with you regarding maintaining the same physical parameters with the alligator leads as during calibration. My conclusion in a nutshell was that the test fixture was easier to maintain consistency due to the nature of having things solidly mounted in place on a PCB. It is a better way to do it.. however.. clip leads do work very well. My impression was that it basically boils down to what a person wishes to do with the results and what kind of accuracy or repeatability one requires.


Regards,
Gregg
--
VE6WGM

You must understand in TDR mode resolution depend from start/stop and points count

So if you look at bottom you can see max range. Resolution max range / points count

Example stop start 10M - 200M and 1001 points on Lite give max ~ 540m, so min step = 540m / 1001 = 0.54m

You must select correct range.

Hi Gregg,

I have used my nanoVNA tweezers successfully for testing toroid inductors by temporarily inserting the inductor leads through tenth inch grid prototype PCB stock. The grid holes hold the toroid leads in place for the squeezing of the tweezer to make solid connection to the coil. The tweezers have also have been successfully tested up to mid VHF for measuring the characteristics of 0805 surface mount components.

Details of how to make a set of tweezers were published in the Nov/Dec, 2021 issue of ARRL's QEX magazine. Some views of the results from testing SMD components can be found within the QEX article support files that can be downloaded from the ARRL QEX files website:



From my experience I have the feeling that as long as you maintain the same lead lengths and spacing of the alligator clips as used when calibrating the nanoVNA using the Open/Short/Load procedure (with a 50 Ohm =<1%precision non-inductive leaded resistor for the load) then the measurements should be reasonably accurate at HF frequencies for leaded components. I would recommend that you try testing some similar value precision leaded components first at the application frequency of interest to confirm that the accuracy is sufficient for your application.

--
73 & Enjoy,
Tom, VA7TA

All the valid points and suggestions (tinyVNA, SDR, cheap receiver), just to add WHY nanonVNA is not a valid solution :)

Those three above "listen" to the signal and show something on screen are create audio that you can hear.

nanoVNA, on the other side, creates the signal and measures what comes back after reflecting from device being tested (1 port), or what passes through tested device (2 port measurement).

nanoVNA will "record" something if you use it in the way you want, but results will be very unpredictable :)

Rich

I did some 2 port measurements on inductors and capacitors in this video. Perhaps this might give you some ideas to get started?



--
VE6WGM

--
VE6WGM

First: for the application in the uBITX the 2 dB dip is no problem. But if you want to try and improve, change the number of turns (up or down) on L5 AND L7 and try tuning the capacitors again. If the dip gets worse, try changing L5 and L7 the other way. This action changes the matching to the filter to optimize it. I have no idea if adding a capacitor in the ground lead is a good idea, but I am almost certain that the filter unit should be well grounded and adding a capacitor will give a path around the filter and ruin the response.

Second: in your application in the uBITX, you don't need the improved steepness that using two units will achieve. I believe the original application for using two units to create a 4 pole filter is in two way FM radios at VHF and UHF frequencies. In this application the radio needs the extra filter performance because this filter is the ONLY filtering and it is setting the ultimate bandwidth of the radio to reject adjacent channel interference. In the uBITX, this filter at 45 MHz is a roofing filter and the ultimate bandwidth filtering happens further along in the signal path at the crystal filter. From reviewing a few data sheets, it seems that when intended for 4 pole filters, they are sold as matched sets, with a marking as noted above for the pins to be connected together in the middle of the filter. Shipping them in matched pairs suggests that they are somehow matched at the factory; I would guess they are matched to be precisely on the same frequency, though the polarity marking suggests to me that something more complex may be involved.

Think i'll go the SDR way as i allready have one. Now just getting/making a directional antenna. My son had brought his world receiver about a month ago and the interference is all over the frequencies. We suspect some converters of solar panels in the neighbourhood.

I read about i think the tinyVNA. Don't know the differences between the two. I bought my nanoVNA to tune the SWR of my antenna. Now i'm discovering other possibilties.

Here's a spreadsheet that shows how the source and RxLO PLLs are programmed to use different harmonics.?? I've also included \the source code from si5351.c that actually implements it.

You can enter a desired frequency in the yellow highlighted cell to see what happens.? That frequency is compared against the threshold (and multiples) to decide which set of harmonics to use (highlighted in green, according to how it works out), and in turn, that determines what the PLLs in the SI5351 are programmed for.

There's a matrix that compares the Source and RxLO frequencies, and all the harmonics, and you can see that only one of them is 5kHz (which is highlighted in red).

You can also change the "threshold" frequency (normally 300 MHz).

On 5/7/22 7:52 AM, Jon MacGahan via groups.io wrote:

I'm just learning about the Nano-VNA; I don't actually own one yet. I understand that the source is a square wave, which contains odd-order harmonics. Suppose I want to measure return loss on some device that has a narrow-band response. It has very good return loss at F but very bad return loss at 3*F, 5*F, etc. Does the Nano-VNA somehow filter the reflected signal so that it only responds to F, not 3*F, etc. I.e., does the Nano-VNA processing include something like a digital filter that eliminates the harmonics? Otherwise the return loss response would be corrupted by the harmonics.

It's not done with digital filters, it's done with clever choice of harmonics for the receiver LO. The analog and digital bandwidth of the receiver is less than 1 kHz.

The transmit signal has fundamental, 3rd, 5th, etc.

The LO is generated so that it's 5 kHz away at a *different* harmonic, i.e., the transmit will be on the 3rd, and the LO will be on the 5th. So the LO harmonics don't line up with the transmit harmonics.

As a result, the NanoVNA receiver doesn't even see the "wrong" harmonics of the source.

Try it out with an excel spreadsheet where you can set the harmonic numbers and Synthesizer frequencies, it's really quite clever.

(Of course I am assuming that I am operating the Nano-VNA at a low frequency - not operating on the harmonics.)

Please disregard my previous post. I found the answer.

I'm just learning about the Nano-VNA; I don't actually own one yet. I understand that the source is a square wave, which contains odd-order harmonics. Suppose I want to measure return loss on some device that has a narrow-band response. It has very good return loss at F but very bad return loss at 3*F, 5*F, etc. Does the Nano-VNA somehow filter the reflected signal so that it only responds to F, not 3*F, etc. I.e., does the Nano-VNA processing include something like a digital filter that eliminates the harmonics? Otherwise the return loss response would be corrupted by the harmonics.

(Of course I am assuming that I am operating the Nano-VNA at a low frequency - not operating on the harmonics.)