The idea
In real-time sampling, the instrument takes all the samples of a capture in a single pass. Its limit is the speed of the converter. In equivalent time, the instrument takes advantage of the fact that the signal repeats identically: it takes a few samples on each repetition, each time at a different point in the cycle, and then places them in their correct position to build a densely populated waveform. The result is an effective time resolution beyond what the converter could achieve in a single pass.
Two variants
- Sequential: after each trigger, the sample is taken with a slightly longer delay than the previous one, and in this way the signal is traversed step by step.
- Random: the instrument samples continuously and measures the interval between the trigger and each sample; because the relationship is random, the samples fall at different positions in the cycle and are put in order afterward. It also makes it possible to see what precedes the trigger.
What it requires
- The signal must be repetitive and stable.
- A very stable trigger, because trigger jitter translates into scatter of the points.
- The instrument must measure the delay between trigger and sample precisely.
What the engineer sees in practice
A waveform made up of many points that gradually fill in as the signal repeats. Horizontal resolution increases, and signals with frequency components higher than the real-time sample rate can be observed, provided the analog bandwidth allows it. Any variation between cycles, such as noise or jitter in the signal itself, appears as a thickening of the trace.
When it does not apply
- Single or sporadic events: since they do not repeat, there is nothing to reconstruct them from.
- Signals that change from one cycle to the next, such as random data with no repetitive pattern.
- When you need to know what happened in a specific event and not an average picture of many cycles.
In those cases, real-time sampling at the highest available rate is the right choice.