VNA signal processing

[xxxNOAHxxx]

How does the VNA do the signal processing .i.e. how it can measure the amplitude and phase of the signal? Does VNA also do the upconversion and downconversion of signals and translate RF to IF to baseband to IQ signals? Does it have same hardware architecture as VST, from RF signal to a series of attenuators/amplifiers then fed to a mixer to downconvert it to IQ signals?

[KittF]

Many good questions.  

As a starting point, make sure you read this document because a picture is worth a 1000 words 

http://www.ni.com/white-paper/11640/en/

There are different kinds of VNA.  NI has 2 types: PXIe-5630 and PXIe-5632.  Since the PXIe-5630 is simpler lets start with that.  It is simpler because it only takes measurements in one direction (i.e. Forward direction).  Specifically, it has a sine wave signal source from port 1, that sweeps across frequencies (and optionally power), while taking relative measurements of what reflects back (S11) and what gets transmitted (S21).  In this case, it must perform vector signal analysis, both magnitude and phase, on the incident, reflected and transmitted signals concurrently.  This means there are essentially 3 VSAs (aka receivers) within a VNA, and yes, they downconvert the RF signals to IF (with diff IF filters and amplifier) and then digitize the signal for magnitude and phase comparison.  In order to measure the signal traveling in a particular direction (vs the opposite), the receivers are connected to directional couplers that siphon a small % of signal in a given direction. All receivers are based off the same LO to optimize their ratio measurements.  

That's the basics.  If you add support for Reverse measurements (ie signal coming from port 2), then you either add some internal switching/routing of the source, or you add a second source and a 4th receiver.  The 5632 dose the latter.  This gives you S12 and S22 measurements. 

I believe that the ratio'd measurements/results happen in the driver layer vs hardware, but I don't really know.

There are a lot of other cools things you can do with a VNA, but that is another post.

Check out these good resources including

VNA Webcasts

http://www.ni.com/white-paper/11641/en/

VNA discussion groups

https://decibel.ni.com/content/groups/network-analysis

The VNA help doc is good (see section called Vector Network Analyzer Architectures. Also look at block diagrams)

http://digital.ni.com/manuals.nsf/websearch/C37B00483716611786257A4E00557F83

[JasonS]

There is a great white paper detailing VNAs and how they work in NI's RF Academy:

http://www.ni.com/rf-academy/measurements/

Have a look at the Introduction to Vector Network Analyzers paper. It goes into great detail about the fundamentals of the measurements that a VNA makes.

[Qiwei6]

Thanks, KittF

The questioner has another question. Is it possible to simulate the VNA measurement method by VST sweeping method, and then measure the S parameter? If using VST, how to determine the phase?

[Wild_man]

Hi Qiwei6,

This is an old thread, and your question seems to be only tangentially related.

I recommend taking a look here: https://forums.ni.com/t5/PXI/Can-VST-measure-VSWR-or-insertion-loss/td-p/3085100

and then creating a new thread if you have follow up questions.

As mentioned in the thread linked, creating VNA-like functionality is possible, but typically requires a lot of effort and external equipment to get similar performance to a real VNA, and is generally not an economical solution.

Thanks,