Is there any practical way to make a VNA that provides the same measurements over the frequency range of ~10 Hz to maybe 1 MHz?

Is that something that could be done with computer or cellphone apps?
I know that you can get FFT and scope apps, but I would like to see the Smith chart for an audio network and other parameters.
Maybe with a variable normalizing impedance....

.... My Christmas wish list.

Neil,
The HP3577A VNA with HP35676A test set covers from 5Hz to 200MHz. Only weighs about 70lb.
Some versions of the NanoVNA-H4 firmware work down to 800Hz (I don't know how well).
I don't know if there is an affordable and portable unit that works below 800Hz.
--John Gord

It's not quite what you want but the old HP 4800A Vector Impedance Meter does 5 Hz to 550 kHz. Manual (as in 'hand operated') sweeping though!
DaveB, NZ

Since the NanoVNA covers from 10kHz up, one can use the sound system of a PC with suitable software to provide the lower frequency coverage. Even the most basic sound chip in a PC will cover 20Hz to 20kHz, and often the low frequency coverage extends to a few Hz. Some better audio cards also offer coverage beyond the audio range, to 40kHz or even more.

Software like ARTA-LIMP allows making useful measurements.

Smith charts, I don't know. I don't use them much myself, as for most situations I find standard X/Y charts more practical. But of course it's perfectly feasible to write VNA software for the PC sound system, that generates Smith charts too.

an oscilloscope would be a much better option. Particularly those that have the audio range.

Otherwise, could divider circuits, say a 10 or 20 to 1 ratio be used to measure audio frequencies? Obviously you'd have to read the NanoVNA with the ratio in mind, but perhaps that is an option one could explore.


Regards

Colin

On Mon, Nov 11, 2024 at 11:11 AM, Neil Preston W0NRP wrote:

Is there any practical way to make a VNA that provides the same measurements
over the frequency range of ~10 Hz to maybe 1 MHz?

I have seen articles describing ham use of, and there are commercial versions of, UPconverter/DOWNconverter products for using low(er)-frequency VNAs to measure higher-frequency and even waveguide components.
All of these, however, use a sample of the analyzer's local oscillator with which to lock the up/downconverter. https://coppermountaintech.com/frequency-extension/ https://www.eravant.com/products/vector-network-analyzer-extenders

Perhaps some clever ham could devise a hack to do the reverse, DOWNconvert the VNA port 1signal with a mixer and after passing the resulting audio frequency signal through the device being tested, UPconvert the AF signal using another mixer and the same LO and feed it to VNA port 2. Likely some filtering and/or image-reject mixers would be needed. I am not certain whether correct phase relationships would be preserved going through the mixers, though.

Another problem would be that the IF filters in the VNA may be too wide for audio use; there may be software workarounds for this.

73, Don N2VGU

On Mon, Nov 11, 2024 at 07:58 PM, Manfred Mornhinweg wrote:

Smith charts, I don't know. I don't use them much myself, as for most
situations I find standard X/Y charts more practical. But of course it's
perfectly feasible to write VNA software for the PC sound system, that
generates Smith charts too.

Only if the PC sound card provides phase information. I do not know whether or not they do.
73, Don N2VGU

There was an article in Jan/Feb 2024 QEX describing using a spectrum analyzer with tracking generator and two mixers, which is essentially the scheme I described above. They used a 30 to 31MHz sweep and a 30MHz LO.
73, Don N2VGU

There have been a couple of audio VNA's (AVNA) in QEX. Here is mine:
http://www.janbob.com/electron/AVNA1/AVNA1.htm
This is not a commercial product, but rather a construction project with a PCB, etc. It covers 10 Hz to 40 kHz with S11, S21 and a bunch more stuff. There is a pretty good pdf instruction book to see how to use it.
The bad news is that the Teensy 3.6 microcontroller/DSP is no longer available. The good news is that there is a redo planned for the project using the Teensy 4.1, that is available. See the groups.io for more:
https://groups.io/g/AVNA1/topic/new_version_of_avna_with/109504643
This is not for everyone as some assembly is required, but the end result is an instrument that is very parallel to RF VNA's.
I believe there is a sound card version of an AVNA around, but I have no personal experience with it.
73, Bob W7PUA

For measurements in the audio-frequency range, you might want to look at the Scimpy software package. It was designed as a means of using a PC sound card, and a simple resistive divider network, to measure the impedance of loudspeaker drivers.

The software calculates and shows and can store the magnitude, and phase, of the speaker impedance (standard .ZMA file format) If I recall correctly, this is enough information to allow the calculation of the resistive and reactive components of the impedance, and hence generate a Smith chart.

This won't help you look above the frequency range your sound card can handle, but it's quite useful within its limitations.

On Tue, Nov 12, 2024 at 01:42 AM, Donald S Brant Jr wrote:

Only if the PC sound card provides phase information. I do not know whether or
not they do.

A basic PC sound interface is simply a group of analog-digital and digital-analog converters. They operate at least in 16 bit resolution, while many have 24 bits. The sample rate is settable, at least to 48kHz, often to 96kHz or more. Usually sampling is coherent between channels, so that accurate phase information can be obtained. The signal/noise ratio usually is somewhere between 80 and 120dB.

The (very inexpensive) sound chip that comes on the mainboard of my computer offers 10 audio channels, which can be configured quite freely as inputs or outputs, with lots of options. A lot of measuring can be done with that, including full VNA functionality. It's just a matter of what the software does.

I have set up my audio chip to provide a stereo line input, a stereo line output, a mono microphone input with separate power supply, and a quadrophonic output for speakers. These four logical audio devices operate completely separate and independently. For example, I can use my line input and output to measure the impedance of an RLC circuit over the audio range with LIMP software, while at the same time listening to music through the speaker outputs with a music player program.

In contrast to desktop computers, the internal audio support of laptop computers tends to be very limited, mainly due to lack of enough audio connectors. But cheap USB audio interfaces are available, and many of them offer a stereo output and a stereo input, so an impedance analyzer can be built around one of those interfaces, using just an extremely simple circuit, which could be as basic as a single resistor and a few connectors. If more input channels are wanted, to make a full two-port VNA with reference sampling, one can use two of those cheap USB audio interfaces, or a single, more expensive one, that has enough inputs in a single unit.

More expensive sound interfaces usually include a DSP chip, differential audio connections, high quality connectors, etc.

Most of the audio measurement software available seems to be centered around speaker and room acoustics measurements, but can also be used in general electronics work. ARTA-LIMP is an old, good, well-known example. Daqarta is also worth looking into, but it's payware whose free features are too limited to be really useful. EasyLCR is a soundcard LCR meter software that works well, but just on a single frequency. And there is a very good program for those who need an audio spectrum analyzer, with a real heap of bells and whistles included: Spectrum Lab, by DL4YHF. There are also many audio oscilloscope programs, but any plain and simple music editing software that allows viewing the waveform can be used as an audio storage oscilloscope with almost unlimited memory. Since I'm old fashioned, I use Cool Edit Pro...

People using the PC sound system for measurements are often at first hindered by Windows mixing input and output audio channels in software, without asking the user. This "feature" can be turned off, but the relevant settings were very well hidden by Microsoft. For those who have stumbled across this problem, here is the recipe, which works fine in my setup of Windows 10:

Right-click on loudspeaker icon in taskbar.
Open sound settings.
At Output, Speakers, choose Device properties.
Choose Additional device properties.
Select Levels.
There you can set up the mix of sources for the speaker outputs. Set to zero all the ones you do not want to directly feed through into the audio output.

My feeling is that with PCs being so common, and almost all of them having audio support, there is little point in building any sophisticated special hardware for audio measurements. Just using an existing PC and suitable software provides high quality measurements, on the cheap. So that's what I use: The PC sound chip in the audio range, and the NanoVNA in the RF range, starting right above audio.

On Tue, Nov 12, 2024 at 09:33 AM, Manfred Mornhinweg wrote:

ARTA-LIMP is an old, good, well-known example.

It appears to be going away:
"March, 5, 2024; Important notice:

The lifetime of ARTA development has come to an end.

After more than twenty years of ARTA software development I decided to stop development and take care of myself.

I have fullfilled my promise that ARTA users will get free update to all versions 1.x.x.....
The ARTALABS web site will be active until February, 2025. The last version and support files will be available for download."

Oops! Those who might want it, get it while you can - or else you will have to ask the Wayback Machine.

More and more creators of well-known software are retiring. Many of them go the extra mile when they do, and put the source code in the public domain, or waive the cost of formerly commercial programs, or at least warn their users to update while they can. Thanks to all of them!

A relatively recent case was Micro-CAP. I had been using the student/hobby version for years, and the end-of-support free release allowed me to get the full version.

When I retire, I intend to release the source code of the (few) programs I have written. So someone else can continue development, if anyone wants.

You don't often see Smith charts at audio frequencies because transmission-line techniques are less common, as is complex impedance matching in general. No reason why they couldn't be supported, of course.

The QuantAsylum QA403 is very nice, hardware-wise, for those who aren't content with sound cards. More expensive than a NanoVNA but still well under four digits.

-- john, KE5FX

One problem with calculations at audio frequencies is that most transmission lines exhibit complex characteristic impedances below around 300 kHz or thereabouts. Telephone cable pairs, if untreated by series inductive loading, exhibit Zo with angles very close to -45 degrees across the voice frequency band.

So the high frequency approximations that lead to convenient expressions for SWR and reflection coefficient are not applicable for most audio applications. Telephone transmission engineers rarely used them, employing instead return loss and reflection loss. Fairly complex charts for determining complex hyperbolic functions were used in the Bell System to make calculations using the exact expressions before computers were widely available.

73,

Maynard
W6PAP

On Wed, Nov 13, 2024 at 10:40 PM, Maynard Wright, P. E., W6PAP wrote:

Fairly complex charts for determining complex hyperbolic functions were used
in the Bell System to make calculations using the exact expressions before
computers were widely available.

One of these was the "Clarke Calculator" invented by Edith Clarke: https://spectrum.ieee.org/edith-clarke-modern-power-distribution
She also designed the turbine system for the Hoover Dam.
73, Don NVGU