We report on an issue regarding the bunch pattern: The bunch pattern read by SCS_RR_UTC/MDL/BUNCH_DECODER shows 48 pulses, but we instead use 50 pulses.

Currently, we have a separate RF window and our 1st pulse Id is 2604. We operate at 2.2 MHz, and as required, 50 pulses. See attachment 1.

This means that the last two pulse Ids are 2700 and 2702. However, these values exceed the number of pulses allowed, and the way the bunch pattern is read (SCS_RR_UTC/TSYS/TIMESERVER:outputBunchPattern,
and the middle layer SCS_RR_UTC/MDL/BUNCH_DECODER) cast the array to 2700 values, so that the number of pulses read is 48 instead of actual 50.

We confirm that we indeed have 50 pulses, as seen by the XGM and by photodiode traces showing 50 peaks.

Since our data analysis strongly relies on the bunchpattern table, this is a critical issue.

For now, the users agreed to use 48 pulses instead of 50 pulses to facilitate the data analysis, but this issue should be investigated.

DOC was called, ITDM is aware and helped finding the reason.

edit 20 Apr 2023, 16:26 : redmine ticket https://redmine.xfel.eu/issues/151949