Vector Network Analyzers' bandwidth

[balnazzar]

I'm posting into this section because it's a beginner question.

Why do VNAs like NanoVNA do have such high starting frequencies? I mean, they start around ~50 KHz and go up to a few GHz.

Why can't they build something that has a different range, like oscilloscopes (e.g. a few Hertzes to some hundreds MHz..).

Thanks.

[radiolistener]

Why do VNAs like NanoVNA do have such high starting frequencies? I mean, they start around ~50 KHz and go up to a few GHz.

Why can't they build something that has a different range, like oscilloscopes (e.g. a few Hertzes to some hundreds MHz..).

few Hertz is not possible on NanoVNA, because it uses si5351 frequency synthesizer. As I remember it's min frequency is about 2-5 kHz. But also RF front-end circuit should support low frequency.

My instance of NanoVNA works good from 10 kHz, but I'm using my own firmware. I tested frequency response of ultrasonic piezo-transducers which working at 24 kHz and can see it's resonant frequency with no issue. In most cases 50 kHz is just software limitation, it can be just hardcoded value or due to specific limitations of algorithms which are used in firmware. I hear reports that some NanoVNA clones has very bad sensitivity at low frequency and almost unusable at 10 kHz. So, in some clones it also limited with RF frontend hardware.

[balnazzar]

Why do VNAs like NanoVNA do have such high starting frequencies? I mean, they start around ~50 KHz and go up to a few GHz.

Why can't they build something that has a different range, like oscilloscopes (e.g. a few Hertzes to some hundreds MHz..).

few Hertz is not possible on NanoVNA, because it uses si5351 frequency synthesizer. As I remember it's min frequency is about 2-5 kHz. But also RF front-end circuit should support low frequency.

My instance of NanoVNA works good from 10 kHz, but I'm using my own firmware. I tested frequency response of ultrasonic piezo-transducers which working at 24 kHz and can see it's resonant frequency with no issue. In most cases 50 kHz is just software limitation, it can be just hardcoded value or due to specific limitations of algorithms which are used in firmware. I hear reports that some NanoVNA clones has very bad sensitivity at low frequency and almost unusable at 10 kHz. So, in some clones it also limited with RF frontend hardware.

NanoVNA starts from 10 KHz, while the V2 starts from 50, but gets up to 3 GHz rather than 1.5...

Mmhh, is there something like NanoVNA, but in the range I'm interested in? Starting from ~100 Hz rather than DC would be OK.

If not, would you share your firmware (I presume for V1) to get it working from 2 KHz?

[radiolistener]

Mmhh, is there something like NanoVNA, but in the range I'm interested in? Starting from ~100 Hz rather than DC would be OK.

for what reason you're needs so low frequency? 

I even cannot imagine what you want to do at that low frequency?

If not, would you share your firmware (I presume for V1) to get it working from 2 KHz?

I don't remember why, but 2 kHz is not possible on NanoVNA. The min stable and usable frequency is 10 kHz.

[PartialDischarge]

I even cannot imagine what you want to do at that low frequency?

There are many reasons why this would be interesting (audio, power line frequency devices...), and VNAs with low frequencies are available but rare (HP 4195A, Anritsu MS4630...) and sometimes a dynamic signal analyser is a good option.

[Bud]

VNAs are designed for and referenced to 50 Ohm or 75 Ohm impedance. There is not much use for it if any use at all at perhaps below 100kHz circuits. At these frequencies you just use an oscilloscope with Frequency Response Analysis feature or an Impedance Analyzer.

[balnazzar]

Replying to all in a single shot:

Indeed I'm after regular FRA, but getting it for cheapo is not a trivial thing (the bare minimum seems to be a 1104x-e with a siglent AWG, for not much less than 900 eur...), so I eyed such little things that cost just 60 eur, hoping they could serve my purpose.

@radiolistener: I was making reference to your post above, where you said that with your firmware, the NanoVNA could get as low as 2-5 KHz..

Another option is a low end Picoscope, but the their awg only gets to 100 KHz (!).

[Bud]

Keysight EDUX 1002G if you can find one.

I am not familiar with nanovna but if it does not have a configurable reference impedance, you are back to 50 Ohm domain. From practical perspective that means the more the measured impedance deviates from 50 Ohm the more the measurement errors are.

[balnazzar]

Keysight EDUX 1002G if you can find one.

I am not familiar with nanovna but if it does not have a configurable reference impedance, you are back to 50 Ohm domain. From practical perspective that means the more the measured impedance deviates from 50 Ohm the more the measurement errors are.

EDUX 1002G? Of course you can find that one, as long as you can shell out the 800 euros required to buy it

Thanks for the 50 Ohm heads-up, I didn't think about it. Was somewhat under the impression that it was selectable like on scopes...

[Bud]

1002G, not 1052G. The first generation of EDUX.

[joeqsmith]

The NanoVNA is AC coupled at various stages and my not like working down to DC.   For lower frequencies, my original NanoVNA seems to perform the best of the low cost VNAs I have looked at. 

If you don't require the phase, maybe a spectrum analyzer and tracking generator would be a better choice.   If you are looking for old products, maybe the HP 3589A would fit your needs.   

[balnazzar]

Thanks! I have just put both the HP 3589A and the 1002G on my ebay notification list...

[joeqsmith]

I own a 3589A if you have any questions about it.  You can measure S11 with it using an external coupler, or they offered a test set for it.   The input can be set to 1Meg and is DC coupled.   Consider it being 30 years old now when considering the price.  The only problem I have had with mine was the battery had to be replaced.     

[balnazzar]

I own a 3589A if you have any questions about it.  You can measure S11 with it using an external coupler, or they offered a test set for it.   The input can be set to 1Meg and is DC coupled.   Consider it being 30 years old now when considering the price.  The only problem I have had with mine was the battery had to be replaced.     

Finding one won't be easy

[Calvin]

Hi,

maybe a soundcard based analyzer (up to 100kHz and 24Bit) might just be what You could use.
Either as internal or as external (USB) Soundcard and running commercial (Arta) or free software (REW, Holm Impulse, etc).
The Quant Asylum A401 to A403 would be a neat little onefitsall soundcard device.
Up to ~10MHz and 14Bit resolution Digilent's Analog Discovery2 could fill the bill.
Offering scope, voltmeter, signal generator functionality as well as VNA, impedance analyzer and some digital functionality (of course with high impedance inputs, no 50R).

regards
Calvin

[jasonRF]

I'm posting into this section because it's a beginner question.

Why do VNAs like NanoVNA do have such high starting frequencies? I mean, they start around ~50 KHz and go up to a few GHz.

Why can't they build something that has a different range, like oscilloscopes (e.g. a few Hertzes to some hundreds MHz..).

Thanks.

VNAs are designed to measure things like S-parameters, which are useful in microwave engineering where you need to take into account the wave-nature of the signals in the electronics.  This becomes important when circuit elements (including wires or transmission lines) are some fraction of a wavelength long.  At 10 kHz, the wavelength is 30 km, so we wouldn’t normally think about S-parameters unless we had a transmission line that was at least 300m long.   

We have a lot of expensive VNAs at work - I would be surprised if any of them went down into the audio band, but of course I have only used them at MHz and GHz frequencies.   Not that I use them that much - I am not at all a hardware engineer - but when we need to do a lot of field testing I take my turn.   

Jason

[balnazzar]

Hi,

maybe a soundcard based analyzer (up to 100kHz and 24Bit) might just be what You could use.
Either as internal or as external (USB) Soundcard and running commercial (Arta) or free software (REW, Holm Impulse, etc).
The Quant Asylum A401 to A403 would be a neat little onefitsall soundcard device.
Up to ~10MHz and 14Bit resolution Digilent's Analog Discovery2 could fill the bill.
Offering scope, voltmeter, signal generator functionality as well as VNA, impedance analyzer and some digital functionality (of course with high impedance inputs, no 50R).

regards
Calvin

Good idea. Some of them get almost to 200 KHz. Will look into those Quant Asylum, thanks!

[balnazzar]

I'm posting into this section because it's a beginner question.

Why do VNAs like NanoVNA do have such high starting frequencies? I mean, they start around ~50 KHz and go up to a few GHz.

Why can't they build something that has a different range, like oscilloscopes (e.g. a few Hertzes to some hundreds MHz..).

Thanks.

VNAs are designed to measure things like S-parameters, which are useful in microwave engineering where you need to take into account the wave-nature of the signals in the electronics.  This becomes important when circuit elements (including wires or transmission lines) are some fraction of a wavelength long.  At 10 kHz, the wavelength is 30 km, so we wouldn’t normally think about S-parameters unless we had a transmission line that was at least 300m long.   

We have a lot of expensive VNAs at work - I would be surprised if any of them went down into the audio band, but of course I have only used them at MHz and GHz frequencies.   Not that I use them that much - I am not at all a hardware engineer - but when we need to do a lot of field testing I take my turn.   

Jason

The strange thing is that a VNA capable of looking into 3 GHz signals is sub-100$, while if you need low frequency FRA you have to spend 10X more..

[jasonRF]

There has been a thread on this recently

https://www.eevblog.com/forum/testgear/cheap-and-small-dynamic-signal-analyzer/msg3189774/#msg3189774

[David Hess]

There are three reasons besides the application:

1 - AC coupling if used limits the low frequency response.
2 - Transformers or directional couplers also limit the low frequency response.
3 - Often the close in phase noise of the oscillator is high making low frequency measurements noisy.

Low frequency VNAs which operate down to DC do exist and are sometimes just called network analyzers.  Nothing prevents implementing an input/output port which operates down to DC, however doing this limits high frequency performance.