While attempting to add the ADC with the HEPI L4Cs into the BSC-ISI model, we discovered that the signals in both models were not the same - that is to say that the signal received on the same channel and on the same ADC was different between the two models. The signals differ by a somewhat constant factor: the transfer function from one signal on the HEPI model to the same signal on the BSC-ISI model is mostly flat, with a light upturn or downturn at much higher frequencies, and with a certain factor difference that is usually a factor of 2.8 or 1/2.8 (that is to say that the signals are highly coherent, but the amplitudes are different, as if the signal from one was scaled by a certain gain relative to the other). Some signals did match, but the transfer function was not EXACTLY one, as it should be for the exact same signal (there were small deviations from 1 at high frequencies).

Since the models are running at different rates, we tried re-compiling the HEPI model at 4k to match the seismic rate. This changed which channels matched  between the two models, but there were still some un-matched channels between the two. 

We are still investigating the source of the problem. In the meantime, the isi2stagemaster model, as well as the individual h2isiitmy and h2hpiitmy models have been checked into the svn should anyone want to look for potential "We've measured the speed of light and it's off by 60 nano-seconds because there's a loose cable" type problems. We've already found one and corrected it, but that certainly doesn't mean there aren't others.