User name brads
Log entry time 10:09:10 on April24,2008
Entry number 234058
This entry is a followup to: 233908
keyword=TDC window change from yesterday looks OKHere are some drift time plots from a recent run (Fig 1) and a run prior to the TDC time window being reduced (Fig 2). The time cut on the histograms has been relaxed so you can see the full window (rather than the 200 ns slice in the regular online gui).
It looks like the change did what it was supposed to. Note the left edge in the randoms is now 500ns earlier. We still have lots of clearance between the left edge and the drift peak.
We could trivially knock another 100--150ns off the window if we wanted. We could also think about pulling 500ns out of the TDC stop and reducing the window another 500ns (followed by recomputing the t0's). Note that none of this affects the tracking performance or analyzer speed (unless we're severely disk-bound). It only saves disk space. I worked out an the drop in average event size for a couple runs before and after the change and it was only a 5% reduction, however the kinematics also changed (changing the background levels) so YMMV.
FIGURE 1
FIGURE 2
