VNA confusion

[antenna]

I wanted to measure amazon ferrite beads so I made a fixture (see picture) where one end connects to the VNA and the other has an SMA to do an OSL calibration. Without a bead, I did the calibration, then replaced the short on the SMA connector. When I set my two traces to read "R and X", and "real and imaginary" (first of all, I thought they were one in the same), I see the "real" shows a point where it transitions from negative to positive value. Second bit of confusion is that the units for real and imaginary are in U. I assume this means I need to normalize by multiplying by 50, but overall I am confused.

Edit, so I am now gathering that these real and imaginary numbers are based on a polar plot and are, in fact, one in the same lol...   I was not expecting, nor was I familiar with, data displayed in that manner. My goal was to see where the inductive, resistive and capacitive regions of that ferrite were, and after remembering the most basics of smith charts, it is clear as day to see where the trace leaves the reactance curve and starts heading off to the right in a resistive direction. Clearly this fixture is not good enough to see the capacitive region, but that was not critical to figuring out I got ripped off on these so-called type 31 ferrites.

Thanks to all who read this, and if I am still missing something, please let me know.

[KasparS]

You may want to consider measuring the set up with the (50 ?) load attached and not the short after calibration as otherwise you will just have all the power that goes through the inductor getting reflected. Which may be why you see different results to those expected. 

[antenna]

The reasoning for the short was to simulate a single loop from the center conductor to ground as if I were measuring just the inductance of the loop with the ferrite added. To replace the short with a 50ohm load is to put that resistance in series with the inductance and I am not sure that would help me characterize the ferrite. I was trying to follow along with this: http://www.reeve.com/Documents/Articles%20Papers/Ferrite%20Beads/Reeve-Hagen_FerriteBeads_P1.pdf.  You may be right, I have no idea.  I suppose then, a better question is, what is the proper way to characterize an RFI supression ferrite (of any ferrite for that matter)?

[jonpaul]

You need an impedance test set.

We use HP 4195A Spectrum/Network analyzer with 41951 Impedance test set and appropriate SMD/Thru Hold adapters to the APC-7 connectors.

Jon

[RoGeorge]

The right side of the chart does not mean resistive.

Fig.6 from page 6 of https://hpmemoryproject.org/an/pdf/an_1287-1.pdf shows how the Smith Chart is made by warping the "normal" but infinite Cartesian Plane of impedances into a finite circle.

Only the horizontal line means resistive behavior.  To the most right side means infinite resistance, to the most left of the horizontal line means zero ohms.  In the middle of the circle is the nominal (calibrated against) value, which is usually 50 ohms.

Everything above that line is inductive, everything below that line is capacitive.

The yellow trace in the pic shows only inductive behavior (because is in the upper half of the chart), which is expected since you are measuring a series piece of wire, which is a series coil, an inductance.

[antenna]

The right side of the chart does not mean resistive.

Only the horizontal line means resistive behavior.  To the most right side means infinite resistance, to the most left of the horizontal line means zero ohms.  In the middle of the circle is the nominal (calibrated against) value, which is usually 50 ohms.

Everything above that line is inductive, everything below that line is capacitive.

The yellow trace in the pic shows only inductive behavior (because is in the upper half of the chart), which is expected since you are measuring a series piece of wire, which is a series coil, an inductance.

When I mentioned moving toward the right, I was referring to crossing more of the constant resistance circles while deviating from the constant admittance curve (which I now realize the latter is irrelevant), but that was definitely the wrong way to look it anyhow. I did explain what I was thinking very poorly.   I've since made a new fixture that allows me to use the shortest wire possible. Now I can see (with other ferrites) that there is a point the parasitic capacitance starts to take over and the inductive reactance starts dropping, eventually reaching zero, then going capacitive in the lower half of the chart.  It was these points I was searching for.  I really wish Siglent would have made an option to view |Z|, R and X in a normal chart rather than a smith chart as that would make the point at which the inductive reactance is highest clearly visible so that I didn't have to slide my marker along the yellow line to find it.

Thanks to everyone who has replied! I've tuned many antenna with my VNA, but testing these ferrites has been a new experience for sure!

[antenna]

Ill make a quick video of what I'm doing, maybe someone can check it out and see if I'm doing this right.......

[antenna]

https://youtu.be/yLm822c-txE

I did call the resistive value "impedance", but caught the mistake a moment later. I wasnt going to retake the video or edit it with a correction. You'll get what Im saying.... so about 71ohm absolute value on the impedance at 27MHz at the end of the video there

[RoGeorge]

The point with maximum inductive reactance seems indeed the one you pointed at minute 4:54, of about \$+j55.94\Omega\$.  From that frequency up, the magnetic properties of the ferrite gland start to decline.

[tautech]

No expert at this but I think you're going about this wrong when instead of trying to characterize the bead single port (S11 reflection) on a piece of wire instead of how it really impacts on your antenna setup.
However to characterize the bead at the frequency range of interest I like jonpaul think it should be on a through fixture for which you can probably get a result on just a SMA coax threaded through the bead and into Port 2 (S21 through) and see how much it moves the ideal 50 ohm center point at the frequency of interest. (marker values)

Anyways if I get a chance in the morning I'll do some tests and pop up some screenshots.

My VNA buddy should chime in later with some better advice.

[antenna]

No expert at this but I think you're going about this wrong when instead of trying to characterize the bead single port (S11 reflection) on a piece of wire instead of how it really impacts on your antenna setup.
However to characterize the bead at the frequency range of interest I like jonpaul think it should be on a through fixture for which you can probably get a result on just a SMA coax threaded through the bead and into Port 2 (S21 through) and see how much it moves the ideal 50 ohm center point at the frequency of interest. (marker values)

Anyways if I get a chance in the morning I'll do some tests and pop up some screenshots.

My VNA buddy should chime in later with some better advice.

Part 2 of the document I was following http://reeve.com/Documents/Articles%20Papers/Ferrite%20Beads/Reeve-Hagen-Poulsen_FerriteBeads_P2.pdf does that, and I was tempted to try, just haven't got around to making a fixture for it yet (as I am out of connectors now, need to order them).  Turns out, there was a 3rd part to that document as well which I have yet to read. http://www.reeve.com/Documents/Articles%20Papers/Ferrite%20Beads/Reeve-Hagen-Poulsen_FerriteBeads_P3.pdf

However, I do not think simply passing a cable through the bead and to the other port is going to work on account of it being a shielded cable. I would need to have a portion of the center conductor exposed, while shielding from the environment (as in part 2 of that document linked above on pg. 6 like they did with the cookie tin).  If I built that, I could just use the spectrum analyzer and tracking generator, normalize the s21 without the bead, then pop in the bead and close the lid.  That would give a very accurate result in dB.  Another issue is that if, say the coax shield on my feed line, is a particular portion of a wavelength and presents a capacitive reactance to the common-mode current, adding inductance in the form of a low quality ferrite can actually make the path easier to follow by cancelling the reactance, so I also need to be able to measure the impedance on the path I want to use the ferrite on to fully get a hold of the common-mode issue (another day lol).  In reality, All this has led me to a new plan, using a 1:1 current balun at the antenna, but all this effort is not to waste as that won't be the only place I use ferrites once I understand their behavior.

[tautech]

Yes after a little research I believe you are on the right track.
Another link that may be worthy of study on the subject of characterizing ferrites:
https://www.kn5l.net/S21adapter/

Also maybe my 315 MHz antenna project log could be of interest:
https://www.eevblog.com/forum/rf-microwave/antenna-project-log/

Some antenna characterization:
https://www.eevblog.com/forum/rf-microwave/what-really-is-this-antenna/

Please keep us posted on your findings.

[antenna]

Yes after a little research I believe you are on the right track.
Another link that may be worthy of study on the subject of characterizing ferrites:
https://www.kn5l.net/S21adapter/

Also maybe my 315 MHz antenna project log could be of interest:
https://www.eevblog.com/forum/rf-microwave/antenna-project-log/

Some antenna characterization:
https://www.eevblog.com/forum/rf-microwave/what-really-is-this-antenna/

Please keep us posted on your findings.

I will. Thank you for the research and links!

[Marsupilami]

Ill make a quick video of what I'm doing, maybe someone can check it out and see if I'm doing this right.......

Sorry if it's not an issue anymore, I'm catching up
I think you picked the point accurately on the video. I wouldn't do this on a switch chart, are you sure your analyzer doesn't have an impedance display?
Anyway on the smith chart take the upward curling curves as your ladder steps and the highest (or most right) point on that ladder is the point of max inductive reactance. You see the Im(z)=1 line is marked with green. This is 50Ohm normalized so that means +50Ohm reactance. In your case it's 55Ohm from which your analyzer even calculates that at 2.8MHz it's 3.1uH.
Ultimately though as far as I know you should care about the highest impedance point that takes resistance into consideration as well and it's hard to visualize on the smith chart.


Regarding one vs two port measurement you're also good, assuming that at the frequencies you're Working at the effect of the ground conductor is negligible.
Eventually it's the same 1 port component you're measuring. (There are subtleties to this, but I don't think any of that is important here.) If you identified the impedance of your 1 port component you can easily calculate what's its response will be in a 2 port short or shunt setup. (Again assuming that the gnd effects are negligible)