Coherent detection should measure the current velocity, averaged over the sound path, with mm/sec precision or better. This depends on the psuedo-random phase modulation of the 200-MHz carrier to resolve the modulo-one-period ambiguity which is inherent in any phase-sensitive detection. This isn't working very well right now.
On May 3, 1996, three of us (ANG, MCE, and RWB) went to the Slough to try out a new transducer driver. This driver is a FET switching device which converts the phase-modulated 200 kHz analog signal to a +/- 15 V square wave, with low output impedance. We wanted to see if increasing the signal size made the measurement of current velocity more precise.
During this visit to the site we observed operation over a range of signal strengths: we used the new driver to increse the signal size by a factor of four; and we decreased the gain with the standard electronics to reduce the signal by a factor of about 2. As far as we could see from watching the signal in real time, the variability of the envelope time did not depend on the size of the signal.
Data taken this day shows some interesting correlations between envelope variability and the other measured variables. We record the rms value of the received signal for each sample. For one direction of signal propagation, variability of the envelope correlated quite well with large signal. To see this correlation, see Phase and Envelope Times and Signal Strength, Outbound Direction.