Main INDEX
Monthly INDEX
PREV
NEXT
User name R. Michaels
Log entry time 18:01:21 on January27,2003
Entry number 93579
keyword=re: synch problems
After getting back in town and having seen the info about
synch problems, here is my advice and prognosis:
===== Advice =======
=> 1. Always run UNBUFF. This is safer and ok for low rates.
=> 2. In the window from which you ran "runcontrol" inter-
actively, watch for errors like "Event type mismatch" from
the event builder. If it happens, restart CODA.
If it still happens, you cannot run this way.
=> 3. In the analysis, use the *REDUNDANT* data
of helicity (it is in fastbus and VME), event time (in each
ROC), and event type (including bit pattern in TDC).
====== prognosis =======
It appears the problem with runs 20548 etc is that ROC3
(fastbus) got out of synch. I'm not sure, but it probably got
ahead of ROC11 (VME). But the "synch" events reported
very few problems, which is strange and contradictory.
Later this evening I'll come in and add the event type to the
datastream for ROC3 and ROC14, etc. Also on R-arm.
I realize this is one redundancy I did not have.