I guess you are talking about the resonance frequency of a certain mode in a cavity; its surprising that you dont mention transmission measurements (there its a bit easier) anyway go into the cartesian display after having done a calibration for S11 in a frequency range around the exspected resonance and then you should see a dip at resonance..for Q value elavaluation we discuss later

Hey Niranjan Thorali ,

In case that you want to switch the FieldFox from frequency domain to time domain you can do that over measurement set-up -> transform -> transform on. This will do an inverse fourier transformation of the return loss trace. It will display Return Loss (Y-axis) versus time (X-axis), instead of Return Loss versus sweeping frequency.

Nevertheless, a time-domain analysis is a very useful tool to observe the effects of mismatch along with a transmission line system, e.g. something that does not match 50 Ohms. When an RF or microwave signal propagates along a transmission line, a portion of the signal reflects back from any discontinuities encountered along that path. Using the time-domain analysis, the location of each discontinuity displays as a function of time along the x-axis on the FieldFox and the amplitude of the reflected signal, or S11 plots along the y-axis. Knowing the propagation velocity along the transmission line allows to scale the time measurement to physical distance.

Now I am not sure how and why you would like to measure the change in resonance versus time. In the case that you just would like to track the change of resonance versus time because you make changes in the system, I would keep the FieldFox in frequency domain and store the resonance frequency values of each frequency sweep with a time stamp and plot resonance versus time in seconds or minutes. Generally, I agree with Fritz Caspers, that a transmission measurements would be easier .

I hope this is useful information,

Best regards,

Patrick

It all depends how fast those drifts or molduations of the resonance frequency occur and which fraction of the 3 dB bandwidth they amount to.

One option is to go on zero span and set the frequency os a say -3 db point and measure the variations opf amplitude (in S12 or also S11) vs time with a fast sweep.The other option which is little konw is to use time domain for CW excitation and see the modulation sidebands on the VNA providet that the IF bandwith on the VNA is big enough to get the fast modulation pass.

I and my colleagues presented measurements of slow drifts in Q as a function of time in this paper (Figure 5):

https://www.researchgate.net/publication/351664850_Low_loss_dielectric_measurements_in_the_frequency_range_1-70_MHz_by_using_a_vector_network_analyser

The above paper used an algorithm given in this report:

Technical Report Q-factor Measurement by using a Vector Network Analyser (NPL...

If monitoring faster changes, you may be able to fit to data for a broad span with maybe 51 points. There is a brief mention of this possibility on page 20 of the report.

You could use reflection or transmission - but transmission is usually easier as good accuracy can be achieved without VNA calibration. In transmission, weak coupling is used, so the effect of couplings and the external circuit (the VNA) is reduced.