Time-domain network analysis and frequency domain network analysis are very similar measurements. The former applies a spectrum of discrete frequencies to the unknown network: a step or impulse incident waveform is applied and an oscilloscope or sampling head captures the reflected and transmitted waveforms. The latter applies a series of discrete frequencies and captures reflected and transmitted amplitude and phase using phase-sensitive (IQ) receivers.

The frequency-domain VNA approach has better dynamic range because the applied power at each frequency can be constant and relatively high, and the receivers can have restricted noise bandwidth.

TDR/TDT can theoretically be quicker because a single step or impulse could give all necessary information. However, the high sampling-time resolution required by this method tends to call for a sequential sampling oscilloscope such as the PicoScope 9311. This captures only one sample point on each cycle of a step or impulse, so only repeating signals can be tested. Even so, our TDR or TDT solution is still slightly quicker than the VNA. See below regarding multiple forward, reverse, transmission, and reflection measurements.

The PicoVNA vector network analyzer and the Pico 9311 TDR/TDT Sampling Oscilloscope solutions are comparable on primary features and performance:

1. In the context of cable or transmission line testing, time-domain measurements have the advantage of
directly showing impedance continuity v time (TDR) or actual pulse response (TDT). Assuming propagation velocity is reasonably well known, a direct interpretation or readout of impedance v distance is possible. Fortunately, the PicoVNA includes a time-domain readout without extra cost, so both solutions can achieve the ideal readout.

2. Where it is necessary to combine measurements with other system elements (measured or simulated) then scattering parameters, Smith charts, etc. tend to be preferred. While S-parameters and time domain
measurements can be interrelated by an FFT, only the PicoVNA supports both output formats.

3. The VNA is limited to 6/8.5 GHz and best effective time resolution of 120/85 ps. Assuming a line propagation velocity of 2c/3, this will resolve impedance v distance along your line to around 24/17 mm [with short/open fault location around five times better). The PicoScope 9311 fast step has a system transition time of around 60 ps, giving it roughly the same time (2 x 60 ps) and distance resolution, giving it roughly the same resolution as the PicoVNA 106.

4. At its best time domain resolution, the PicoVNA can test at 4096 k discrete sample points. This translates to a maximum path length of around 100 m. At lower resolution, longer path lengths can be accommodated. The longest path length supported by the PicoScope 9311 is theoretically limited by the widest available pulse width of 4 µs, which translates to 400 m in a reflection measurement. In practice, however, at this length, the resolution of a TDR measurement will be lost to high-frequency cable losses and a VNA may match or even outperform it. TDR distance can be increased by instead using the lower frequency capability of PG900 fast step generators.

The larger differentiators between the two are:

1. The VNA automatically sequences reflection and transmission measurements and can do this in both
directions through an unknown 2-port network. Greater speed and much-reduced manual intervention are significant benefits when more than a single port reflection or a single transmission measurement is required. Single-direction measurement is typically sufficient for transmission line measurements, but for devices with transmission losses or gain, bidirectional measurements are far more commonly needed and the automated VNA wins even more convincingly.

2. In differential line applications the PicoScope 9311 is likely to win. It has differential de-skewable step
sources and its two channels can be used to receive either differential reflection or differential transmission. In the case of a differential line without a nearby ground, such as a twisted pair in free space, the PicoScope 9311 can determine differential impedance v distance from a single measurement setup. The differential test Convert web pages and HTML files to PDF in your applications with the Pdfcrowd HTML to PDF API Printed with Pdfcrowd.com stimulus develops a virtual ground (or signal null) at the midpoint between the conductors. The VNA can only stimulate one conductor with respect to the other and cannot resolve differential impedance. It could, however, determine an impedance for a given core and how that changes with length, and could be used in comparison with a gold standard—for instance, both conductors separately measured, the other grounded—in this case.

If we now consider the opposite extreme example of a differential line: the individually screened twinax pair. Here a ground surrounds and isolates the two lines of the differential pair. In this case, the two differential cores can each be measured separately as individual coaxial lines. Their differential impedance will be the sum of the individual impedances. The PicoScope 9311 and the VNA can both do this in a single measurement setup. The two ports of the VNA can be used to separately measure the two cores, but only one TD readout is permitted at a time.

Of course, a great many differential lines lie somewhere between these two extremes with varying degrees of ground and pair coupling. The usefulness and accuracy of the VNA measurement vary accordingly.

Pico is uniquely positioned to support either solution and will be happy to support your decision-making for a given application.