So what I have at hand is a cheap network analyser that can present the measured impedance in form of the values of different equivalent circuit topologies, e.g. R-L series or RL parallel.

What I want to do is measure the electrical parameters of a simple PCB antenna structure, which is typically modeled as as capacitor (mainly inter-winding capacitance) parallel to ( R + L).

So I did the "SOL" calibration and measured S11. Concerning R and L, I chose the available Network Analyser Representation of the equivalent circuit called "RL" (I have no C//(R+L) representation available) This measurement shows the values for both, R-L series and R-L parallel equivalent circuits and I get reasonable values so no problem here.

But how can I determine the value of the capacitor? Is it possible at all? I thought of writing down the formula for the real and imaginary parts of the impedance of the topology I want to model (i.e. C//(R+L)) and solve it using the measured values of S11. Besides of being inconvenient I think I would get a wrong result since any physically capacitance present in the structure is already indirectly included in the values i get for R-L (by having a slightly smaller L than without cap), right?

(Of course the inclusion of the capacitor in the presented R-L values is possible only for a specific range of frequencies because above some frequency the capacitor will dominate and "get visible" in the measurements, but that does not help either...)

  • \$\begingroup\$ What kind of network analyzer do you have? Any Link? How did you attach the analyzer, what kind of fixture did you use? \$\endgroup\$ – Ale..chenski Oct 17 '16 at 19:58
  • \$\begingroup\$ The device is a DG8SAQ found here: sdr-kits.net and I simply connected the TX output of the device directly to the antenna (as I wrote above, before I did the open-short-load calibration). So no special fixture, just TX output of VNA per SMA cable to antenna feed point. \$\endgroup\$ – Junius Oct 17 '16 at 20:03

Set up your network analyzer to plot the response as magnitude of impedance (\$\left|Z_L\right|\$), with log scale on the y axis.

Assuming 1. your network analyzer can reach a high enough frequency, and 2. the connection from the reference plane to the termination you're trying to measure is short, you should see \$|Z_L|\$ increasing from low frequencies up to some peak and then decreasing after the peak.

The equation for the part of the curve where it's increasing is approximately \$Z_L(f)=j2\pi{}fL\$. The equation for the part of the curve where it's decreasing is approximately \$Z_L(f)=\frac{1}{j2\pi{}fC}\$. By picking out a couple of points on the appropriate parts of the curve, you should easily be able to estimate both \$L\$ and \$C\$.

If you already know \$L\$ you can get \$C\$ really easily knowing the resonance frequency (the frequency where the graph reaches its peak) is \$f_R=\frac{1}{2\pi\sqrt{LC}}\$.

  • \$\begingroup\$ I am aware of how inductors and capacitors behave over frequency, but I think just "estimating" is a bit of a problem because I assume the capacitance is in the range of 10-15pF. Thanks anyhow, I will have a look at the "plain" abs(Z) data and see If i get the expected graph. \$\endgroup\$ – Junius Oct 18 '16 at 17:36
  • \$\begingroup\$ If you don't get the expected graph, then your model (L and C in parallel) is not a good match for your actual load, and you need to find a better model. \$\endgroup\$ – The Photon Oct 18 '16 at 17:40
  • \$\begingroup\$ Also, estimating is all you can do because no measurement is perfect. \$\endgroup\$ – The Photon Oct 18 '16 at 17:41
  • \$\begingroup\$ Well, the application notes I found all use this model for that kind of simple PCB antenna so I guess it´s quite realistic ;-) Btw: Any hints on my question on whether the calculation I suggested, is a valid approach or Is there some mistake in it? EDIT: I mean the calculation of C out of the S11 data of the R-L equivalent circuit. \$\endgroup\$ – Junius Oct 18 '16 at 17:44
  • \$\begingroup\$ Sorry, what you said about "Network Analyzer Representation" doesn't make any sense to me. That's a feature specific to the particular VNA you are using, so you'd need to talk to someone who also uses the same model VNA to get any more information about it. \$\endgroup\$ – The Photon Oct 18 '16 at 17:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.