Jan 1 2009 11:22AM | Permalink |Comments (2) |
NASA Tech Briefs has an interesting article about improving time domain reflectometry (TDR). By varying the pulse widths you send down the cable or trace, you can improve the sensitivity of the measurement. You use an autocorrelation function to dig the reflected signals out of the noise. This makes perfect sense. The rising and falling pulses make a pair. If the time between the two edges is always the same, then there might be deterministic reflections from other features that obliterate the reflections of interest. By varying the pulse widths you are sweeping those reflections around this noise and thereby getting a signal with more information, which you can then extract with the autocorrelation. You can also check out my article about TDR.
Related entries in: Analog |