...
9:03 beam change, so no beam for ESA. Beam is expected to be back after ~15mins
stop at 9:24.
Dec
...
/13 Night shift
Zijun, SD
T578 HGTD period starts today but they are not ready to run until Saturday. Some additional time for ITk at the mean time.
...
Whole shift of test runs debugging the loss of event ID match between Caladium and RD53A readout. Some runs got good MIMOSA vs RD53A X,Y correlation at start at least and other runs were not aligned. There were some odd extra event looking like real distinctive RD53A beam hit data at DUT readout but never found a matching MIMOSA event data. TLU ch-2 1 RJ45 port is somewhat suspicious and sometimes could not communicate with FEB. Moved FEB link to TLU ch 32. More stable communication to FEB, but the issue of extra event on RD53A side remains.
Dec
...
/14 Night shift
Day shift: UCSC T788 setup continues during MD. One main issue to resolve is the dead Trg-10 for Caladium. T578 needs both Trg-10 (NIM for Caladium) and Trg-11 (TTL for HGTD). Didn't manage with connect with Doug to debug the dead Trg-10, but he passed on a schematic for the ESTB trigged patch scheme:
...
The issue with Trg-10 is not obvious: EVG rear expansion board ch 11 (LVTTL?) is one of the inputs (the other input is EVG Front trigger 2) to the Lecroy level adaptor in compl mode tuning that into a 1.2V and 300ns width NIM. The only part looking suspicious is the LEMO cable taking the NIM output to J10 looks very tight. Replaced that part with a slightly longer cable. This did effectively bring the Trg10 NIM signal back to Caladium rack but signal shape weak. Decided to connect J10 directly to EVG rear 11 to send TTL signal over long path to tunenel and used TTL->NIM at caladium rack for a much more solid NIM signal to TLU.
Night shift: ZiJun,SD
Finally, realized that TLU RJ45 ch1 defect is one of the reasons for events synchronization failure between MIMOSA and RD53a.
1) FEB can not get trig signal from TLU ch1;
2) MIMOSA can receive trigger signal from TLU ch1, but the TLU words from TLU ch1 are wrong. Left plot is the TLU words distribution from ch1, while right plot is the TLU words from ch0.
Now, connect TLU ch0 to MIMOSA, and ch2 to FEB.
5:26 am New version of YARRproducer stabilized the synchronization between MIMOSA and RD53a readout. Previous version used local event counters with no direct check of event ID to confirm synchronisity. It was seen that block of e.g. 1000 events are offset by a few events between MIMOSA and RD53a, and then this offset would slip every ~1000 events so that sometimes by chance the synchronization is achieved by luck and then slips off again after a while. New version defies the RD53a event number using the TLU distributed event number (offset by 1) to automatically achieve synchronization.
...