Hi Jordi, I had this problem about a year wanting to buy a spectrum analyser. Well technology has come to the rescue. I bought a TinySAUltra for around $260 Australian and a TinyNano Vector analyzer for $60 Australian and I am just staggered how good they are. Sure they have limitations compared to the bench top units but value for money they are awesome. Normally I am a luddite when it comes to computer technology but when it comes to this kind of technology I can be a technophile.
More details in future videos, moving the cal plane down to jig/circuit interface levels with all kinds of connects and cavity filter jigs, etc. Equations if necessary (simpler ones at first then build up complexity). Accurate measurements is a huge black arts problem.
Agree completely. VNA calibration is a *huge* topic with lots of room for additional presentations / videos. I named this presentation "VNA Calibration BASICS" because it only covers the most fundamental concepts at a relatively high level. If you'd like something a little more in-depth in the meantime, you can download our "Fundamentals of Network Analysis Primer" here: www.rohde-schwarz-usa.com/VNA-Fundamentals.html. (Don't let the name fool you, it's actually got quite a bit of math in it :))
As others have pointed out it would be nice to have a video explaining port extension and electrical delay and verifying cal plane is at the input of dut.
More detailed videos on real world applications would be nice. Rarely are people encountering the connector types on the VNA or even the same impedance's. So how do you properly calibrate a system that goes from 50 ohm N type VNA to a 75 ohm F type. How do you handle this when you need to jump from a N-type to an SMA to a 50-75 ohm F-type pad?
@@Floyd.B Agree. VNA Calibration is such a huge topic and that's why I called this video "Understanding VNA Calibration BASICS" :) I'm planning some additional content on calibration in the not-too-distant future, so thanks for the suggestion!
That's a good, but somewhat involved question :) Calibration is concerned with removing so-called "error terms" and for a one port measurement these are reflection tracking, directivity, and source match. This in turn requires three cal standards: open, short, and match (or load). Since one port normalization uses only on standard (an open or a load), it cannot remove all three error terms. Normalization corrects for reflection tracking (only), but the full calibration removes all three terms. Hope that helps!
At a minimum, you should calibrate any time something *substantial* has changed since your last calibration. Obviously, if you're using different cables, connectors, you would want to run a new calibration. If the environment changes (e.g. moved the VNA to a different lab), then you really ought to run a new calibration as well. Many people will run new calibration at the start of each testing session or before "important" measurements. For example, if I'm running an S21 measurement to figure out the rough passband of an unlabeled "mystery filter" I found in an old box somewhere, I probably wouldn't be terribly worried about running a new cal. On the other hand, if I were adjusting a filter or if I were designing/testing a filter, then a valid calibration is much more important. One of the reasons that autocalibration units (like the one I mentioned in the video) are so popular is because they allow you to run a full cal very quickly -- if it only takes a couple of button pushes and a minute or so to get a fresh cal, why not? :) Normally a network analyzer is calibrated over a range of frequencies, so as long as you stay within that frequency range you shouldn't need to recalibrate.
In my experience, the physical standards are always labeled "thru" (see the pictures on the first slide and on slide 12), so that's how I labeled the graphic at 9:26. I've seen it both ways in writing, but I prefer the normal spelling of "through" and this is also the spelling we use in our "Fundamentals of Network Analysis" book and whitepaper: www.rohde-schwarz-usa.com/VNA-Fundamentals.html.
Wow. Apparently Rhode Schwarz is a manufacturer but these videos are as good as that of educator's. That's really awesome. Keep up the good work.
Thanks! Education is an important part of our mission at Rohde & Schwarz. Stay tuned for more videos coming soon!
I love all those RF things! Unfortunately, only the calibration kit costs more than my house probably.
🤩🤩🤩
Hi Jordi, I had this problem about a year wanting to buy a spectrum analyser. Well technology has come to the rescue. I bought a TinySAUltra for around $260 Australian and a TinyNano Vector analyzer for $60 Australian and I am just staggered how good they are. Sure they have limitations compared to the bench top units but value for money they are awesome. Normally I am a luddite when it comes to computer technology but when it comes to this kind of technology I can be a technophile.
More details in future videos, moving the cal plane down to jig/circuit interface levels with all kinds of connects and cavity filter jigs, etc.
Equations if necessary (simpler ones at first then build up complexity). Accurate measurements is a huge black arts problem.
Agree completely. VNA calibration is a *huge* topic with lots of room for additional presentations / videos. I named this presentation "VNA Calibration BASICS" because it only covers the most fundamental concepts at a relatively high level. If you'd like something a little more in-depth in the meantime, you can download our "Fundamentals of Network Analysis Primer" here: www.rohde-schwarz-usa.com/VNA-Fundamentals.html. (Don't let the name fool you, it's actually got quite a bit of math in it :))
As others have pointed out it would be nice to have a video explaining port extension and electrical delay and verifying cal plane is at the input of dut.
I actually am working on a presentation "Understanding De-embedding" that will cover these topics -- should be ready in a few weeks
th-cam.com/video/LW5yV5qmnAg/w-d-xo.html
Great video!
More detailed videos on real world applications would be nice. Rarely are people encountering the connector types on the VNA or even the same impedance's. So how do you properly calibrate a system that goes from 50 ohm N type VNA to a 75 ohm F type. How do you handle this when you need to jump from a N-type to an SMA to a 50-75 ohm F-type pad?
Those are great suggestions. I'm working on more VNA related content and will add those ideas to the queue. Thanks for the ideas!
@@Floyd.B Agree. VNA Calibration is such a huge topic and that's why I called this video "Understanding VNA Calibration BASICS" :) I'm planning some additional content on calibration in the not-too-distant future, so thanks for the suggestion!
Regarding one port calibration: What kind of an error source is fixed by Full calibration that is not fixed by Normalization?
That's a good, but somewhat involved question :) Calibration is concerned with removing so-called "error terms" and for a one port measurement these are reflection tracking, directivity, and source match. This in turn requires three cal standards: open, short, and match (or load). Since one port normalization uses only on standard (an open or a load), it cannot remove all three error terms. Normalization corrects for reflection tracking (only), but the full calibration removes all three terms. Hope that helps!
Can we leave the port open in the air instead of using the kit for Open calibration when we are performing SOLT calibration?
You could, but it's safer to use a cal standard - less chance of external signals entering the VNA port and influencing the calibration.
1. Why changing the frequency void the calibration? 2. If do we need to calibrate each time we use the VNA?
At a minimum, you should calibrate any time something *substantial* has changed since your last calibration. Obviously, if you're using different cables, connectors, you would want to run a new calibration. If the environment changes (e.g. moved the VNA to a different lab), then you really ought to run a new calibration as well. Many people will run new calibration at the start of each testing session or before "important" measurements. For example, if I'm running an S21 measurement to figure out the rough passband of an unlabeled "mystery filter" I found in an old box somewhere, I probably wouldn't be terribly worried about running a new cal. On the other hand, if I were adjusting a filter or if I were designing/testing a filter, then a valid calibration is much more important. One of the reasons that autocalibration units (like the one I mentioned in the video) are so popular is because they allow you to run a full cal very quickly -- if it only takes a couple of button pushes and a minute or so to get a fresh cal, why not? :)
Normally a network analyzer is calibrated over a range of frequencies, so as long as you stay within that frequency range you shouldn't need to recalibrate.
in the beginning of the video THROUGH VS 9.26 THRU . Hmmmm
In my experience, the physical standards are always labeled "thru" (see the pictures on the first slide and on slide 12), so that's how I labeled the graphic at 9:26. I've seen it both ways in writing, but I prefer the normal spelling of "through" and this is also the spelling we use in our "Fundamentals of Network Analysis" book and whitepaper: www.rohde-schwarz-usa.com/VNA-Fundamentals.html.