1. Home
  2. Electronics Design
  3. Vector Network Analysis

Vector Network Analysis

Gentlemen, Ladies, Transexuals and Australians,

Do we *really* need S22 and S12 any more? I know they're nice to have

- and indeed essential for some designers - but increasingly the trend in 'mini-VNAs seems to be to dispense with those measurements and focus exclusively on S11 and S21. And if we're honest about it, those last two S parameters are where 90% of the action is. Now don't get me wrong! I have a lab grade HP VNA that probably cost the thick end of 100k quid back in the 90s and it still works great up to 6Ghz. HOWEVER, it's HUGE, MASSIVELY HEAVY and highly complex to use. The new generation of pocketable mini-VNAs are carving their own niche in the market and must have the big dinosaur manufacturers rattled (and if not, they jolly well should be). These new kids on the block are intuitive to use, extremely cheap to buy and perform the 90% of 'the stuff you need' to measure, perhaps not to lab grade, but 'good enough' for initial development. And they don't take half a kilowatt to run, either. So, my question is - are S22 and S12 dead?

Example:

formatting link

Reply to
Cursitor Doom
Loading thread data ...

Sometimes, in automatic setups, it's essential. There have been times I wished for a four-port VNA, so that I could measure all 16 responses from S11 to S44 without juggling cables. A manufacturer of large numbers of four-port gadgets --which I wasn't-- would certainly need such a system. I just moved cables around in those few cases I had to deal with it, but it certainly was a nuisance and a cause of wear and tear on the hardware.

For occasional lab use, just S21 and S11 are good enough, I agree. I don't think the mini-VNAs get anywhere near the performance of the HP boat anchors, yet.

Jeroen Belleman

Reply to
Jeroen Belleman

Yes. I have a HP boat anchor and would love to save the space it takes. I haven't yet seen a miniature VNA that has comparable performance, but maybe this will change in the near future.. John

Reply to
John Walliker

Unless you have transmitters in your vicinity, or need to prevent oscillation, or to unilateralize an amplifier, or combine your amp with a mixer or a filter.

Cheers

Phil Hobbs

Reply to
Phil Hobbs

Also if you are doing proper calibration then you need to measure the raw values of all 4 parameters to do the correction, regardless of whether you want the corrected values or not.

The LibreVNA is interesting. There are some isolation problems in the hardware yet to be solved, but it's a very good start and the developer is very helpful.

Reply to
Chris Jones

Do you mean to say that forward and reverse parameters are measured simultaneously rather than sequentially?

Jeroen Belleman

Reply to
Jeroen Belleman

No, as you'd know, a conventional 2-port VNA will measure raw S11 and raw S21 simultaneously with the source of port 1 on, then it will measure raw S22 and raw S12 simultaneously with the source on port 2 turned on.

I just meant that even if you only care about the corrected S11 and corrected S21, in addition to raw S11 and raw S21 it is necessary to measure raw S22 and raw S12 in order to do the correction. And once you have measured them, they are available for display at no added cost.

If you don't bother with calibration, or only do a response calibration, then the above may not apply.

Reply to
Chris Jones

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.