Here's a comparison of runs 13024001, 13027029 and 13027045.  I regerated all the following pdf files now, so that the software used for all is up to date.  Note: run the earlier two runs were taken with both rdos having 7 time bins, while the last run has one rdo with 7 and one rdo with 5.  I used the first 1000 events, instead of the usual 2000, as the shortest run was just over 1000 events.  All plots are use data only from timebin 2 (counting from 0).

There seems to be sizable differences between the 24th and 27th (additional APVs on octant 3AL are dead), though both runs taken on the 27th seem quite similar, despite them having different time bin configurations.  While the differences between the 24th and 27th are unexpected, this at least verifies at a cursary level, that the data stream seems to work equally well for either number of time bins.