The question of time vs. frequency, as the most useful measurement domain, has long been a controversial topic. In some cases, it leads to rather heated discussions. The argument in favor of a vector network analyzer (VNA), a frequency domain instrument, is that the dynamic range and signal to noise ratio (SNR) of a VNA are much better than they are for a time domain reflectometer (TDR). The argument in favor of TDR measurements is that they tend to be lower cost and are taken from a direct reading, so there is little to interpret. Fortunately, most new TDRs can also transform measurements to S-parameters (much like a VNA) and most new VNAs can also transform to time (providing TDR equivalent data).
Having said all of this, the measurement of a PCB plane using a VNA may not be as straightforward as you might expect. One simple trick makes it easy.
A 2-sided 63mil FR4 PCB is used as an example in this article. The PCB is measured using both a VNA and a TDR. The results are compared, the error source is identified, and the trick is employed to correct the error.
The PCB is shown in Figure 1 connected to a SD-24 20GHz/17.5ps sampling head installed in a Tektronix 11801B DSO mainframe. The board is connected to the instrument via an SMA edge connector and a SMA male-SMA-male barrel adapter. The instrument cursor table is enlarged and shows a characteristic impedance of approximately 3.36 ohms or 10.4dB ohms. I say approximately because the cursor resolution on this instrument is somewhat coarse, but it is pretty close to 3.36 ohms.