Getting a clean signal to your oscilloscope has become a complicated business. How much does probe loading matter? How does the probe's calibration affect the signal? Are there reflections to worry about? If you're de-embedding a fixture or channel, how do you separate de-embedding problems from signal problems? In this paper, practical examples are used to show how these considerations affect high-speed measurements, and how to apply knowledge about the tools at hand to ensure the best results. Case studies include illustrations of different types of probe correction, choosing the cut-off frequency for a de-embedding filter to best balance signal and noise, and removing reflections from a probed LPDDR2 measurement when S-parameter models for de-embedding are not available.
展开▼
