Open-short Normalization Method for a Quick Defect Identification in Branched Traces with High-resolution Time-domain Reflectometry

Abstract

Time-domain reflectometry (TDR) that employs electro-optical sampling affords excellent resolution at the femtosecond level and exhibits a comprehensible impulse waveform, thereby allowing quick defect identification in a single trace. However, it remains challenging to identify a defect in a trace of multiple branches; the TDR waveform is complex. Generally, the TDR waveform of a defective unit features defect-dependent reflection (DDR) and defect-independent reflection (DIR). DDR is contributed by a branch with the defect; DIR is contributed by the remaining good branches. The DDR (not the DIR) is required to analyze the defect; however, the DIR tends to overwhelm the waveform, rendering interpretation difficult. In this work, we use an open-short normalization (OSN) method to eliminate the DIR. The resulting DDR immediately identifies the defect location and type. The OSN method was verified using both simulation and measurements.