Application of e-TDR to achieve precise time synchronization and controlled asynchronization of remotely located signals
MetadataShow full item record
Time Domain Reflectometer (TDR) measures the electrical length of a cable from the applied end to the location of an impedance change. An impedance change causes a portion of the applied signal to reflect back based on the value of its reflection coefficient. The time of flight (TOF) between the applied and reflected wave is computed and multiplied with previously determined signal propagation velocity to determine the location of the impedance change. We intentionally open terminate the output end of the cable which makes the reflection coefficient be maximum (=1) to measure its electrical length. Conventional TDRs designed for testing integrity of long cables use various closed pulse shaped test signals i.e. the half sine wave and the Gaussian pulse, that disperse (change shape) and change velocity while propagation along the cable. Quoting Dr. Leon Brillouin’s comments on electromagnetic energy propagation , “in a vacuum, all waves (e.g. frequencies) propagate at the same velocity, hence withoutdistortion, whereas in a dispersive lossy media, except for an infinitely long sinusoidal waveform, distortion will occur due to frequency dependent velocity.” This signal distortion generally degrades the accuracy of the measurement of the signal’s TOF. We discuss here an Enhanced Resolution Time Domain Reflectometer (e-TDR). The enhanced resolution is due to a newly discovered signal called SPEEDY DELIVERY (SD) by Dr. Robert Flake at The University of Texas at Austin (US PATENT 6,441,695 B1 issued in August 27, 2002). This SD signal has a propagation velocity that is a programmable constant and this signal preserves its shape during propagation through dispersive lossy media (DLM). This signal behavior allows us to use ‘e-TDR’ in applications where remotely located signals need to be synchronized or asynchronized precisely. Potential applications include signal based synchronization of devices like sensors connected in a network. Since the cable carrying data from sensors at discrete and remote locations to a collecting center have different electrical lengths, it is necessary to precisely offset the timestamp of the incoming signal from these sensors to allow accurate data fusion. Our prototype is capable of synchronizing signals 1,200 ft (~ 400 m) apart with sub-nanosecond resolution.