Waveform missing mechanisms and a countermeasure in a random sampling system

This paper reports on a newly revealed waveform missing mechanism and a rational countermeasure in a random sampling system. Random sampling is an equivalent sampling technique which measures waveforms using a finer time resolution compared to the sampling period. Random sampling is usually applied for digital storage oscilloscopes (DSOs). Its design does not require delay lines for introducing the time shift for the measured signals, and theoretically, it is able to measure very-wide-band signals. However, partial waveform missing phenomenon will occur if interference between the trigger signals and hold-off (HO=trigger-enable) signals exists. One of the mechanisms already reported for the phenomenon is where the trigger periods are shorter than those of the sampling periods. Some dithers were added to hold-off signals to eliminate the interference. We call this method the random hold-off method (RHM). In this paper, we introduce the second mechanism where the trigger signals are not evenly divided by hold-off signals. Unexpectedly, RHM has a good effect on the second mechanism. However, we propose a rational countermeasure with no dither to avoid both mechanisms. Our proposed countermeasure controls the timing of the hold-off signals using the data which was measured during a series of acquisitions. We call it the constant hold-off method (CHM). CHM makes it possible to design a low-display-noise and wide-band DSOs because dither produces intermodulation. In a simulation, the waveform acquisition time of CHM is about two-thirds that of RHM