At the last VU-ANL meeting, Hal suggested something to check in the meantime as we locate the bunch-crossing bug. Since we expect single-beam backgrounds to be largely localized in the EEMC, we can check the tower locations for abort gap events. If a particular area of the detector is illuminated as opposed to a general scatter, then we perhaps have our signature for single-beam backgrounds.

For this study, I separate events into four classes: BBC-coincidence, No-BBC-coincidence, BBC-coincidence (abort gap), No-BBC-coincidence (abort gap). For a BBC-coincidence I require 10 ADC counts in a BBC tile. By "abort gap" I mean specifically the blue-beam abort gap (bunch-crossing ID's 111-119). To work in an environment where we can study the effects of turning on and off the coincidence requirement, I have run over the full Run-6 transverse dataset for eemc-http triggers.

Figure 1

In Fig. 1, we see the familiar plot from the previous single-beam study, now, with the full statistics. On the left I show the normalized to sum the entries to unity for both coincidence and non-coincidence events. On the right, I show the comparison for raw number of counts. The blue-beam abort gap is particularly dirty for no-coincidence events.

Figure 2

In Fig. 2, I examine the ADC for the event's high tower. The distributions seem fairly similar between coincidence and non-coincidence events.

Figure 3

Finally, in Fig. 3, one sees a rather stark contrast between the location of high towers in coincidence and non-coincidence events. In the coincidence class, the high towers are fairly evenly spread across the η and φ bins. In the non-coincidence events, including in the abort gap, there is a fairly localized region of towers illuminated. This fits the profile for single-beam backgrounds. My memory is fuzzy on the location in space corresponding to bin numbers. Thus, it reamains to be seen if this profile is consistent with what we expect for single-beam backgrounds in Run-6.