The tagger-OR began the run hampered by significant dead time (
90 ns).
This was traced to improper gate settings in the logic, which were corrected.
The dead time now stands at 10 ns, and, although the effect on the
experiment will be small, it is worth investigating whether the widths of the
gates, and of the re-discriminated, OR signal can be reduced further.
By the end of the run, tagger TDC distributions looked reasonable, with the spectrum of left side times (the side included in our trigger) narrower than the right side spectrum. Among other good things, this indicates that the tagger is setting the trigger timing, as we had hoped. There are occasions when either the E or G RPD counters are setting this timing, and this requires some investigation, although the G-counters, unfortunately, never made it into the TDCs.
Rates in the tagger went as expected, at least if we look at the distributions from our scalers, as opposed to those of CLAS. A simulation indicates that at 105 nA with a 10-4 radiator, the tagger rate in the upper 60% of the focal plane (3.5 Gev to 5.7 GeV) is between 30 and 35 MHz, with the lowest energy paddle being the limit (at 3 MHz). So we can certainly get some of our rate back by increasing the focal plane (the acceptance is still reasonably good at 3.5 GeV) should other factors limit rates. Additions to the electronics should be minimal: presumably just another quad coincidence unit and some cables.