Data summary plots from 12 hours on 7/29/01 (2.31 am to 2.35 pm). In the first two graphs, data from the simple and filter triggers are plotted simultaneously. The simple trigger was disabled so its data appear as a (white) constant. There’s a lag time between the peaks in the event rate and the peaks in the threshold. Overflow count is the number of times the ADC buffer overflowed because data processing fell behind data acquisition. This occurred once, when the event rate was very high (500 evts/min compared to 60 evts/min target). At this time the threshold was climbing as fast as it could to lower the event rate (threshold step was 0.001/minute). So I think the adaptive threshold is useful in this way: we effectively take more data (lower threshold) when it is quiet, and less data (higher threshold) when it is noisy. The fraction of events captured during quiet periods is higher with the adaptive threshold than it would be with a constant one.

A signal of the shape we expect from UHECR?