On July 19th, the Inter-ERP timing manager failed, causing most of the
Inter-ERP TDC data to be overflow. On July 20th, the Inter-ERP was removed
from acquisition and it was returned to acquisition on July 22nd. The
problem is mostly fixed, but there may still be some problems with the data
as described below.
Due to a failure in the ACTEL chip, the normal TDC start for uVAX2 is
not formed. We are currently trying to replace this chip, since no spare
chip is immediately available. The TDC start for a given uVAX is formed
whenever there is a 2/8 coincidence between any of the 8 regular ERP stop
signals and also when there is an ERP muon trigger (EH) on the uVAX. The fix
implemented generates the start for uVAX2 by forming the _OR_ of the start
for uVAX1 with the EH for uVAX2. This fix will produce a start for the
uVAX2 TDC on all real inter-supermodule events, and the data should be
correct in virtually all cases.
The one problem with this fix is that it produces extra starts for the
uVAX2 TDC on some uVAX1 only events, and the uVAX2 TDC will not be cleared
for these events, with the possibility of bad TDC data on the subsequent
event. In this case, all but one of the TDC channels on uVAX2 will be
overflow, so there is no danger of mimicking an upward muon. There is a
clear sent to all TDCs whenever there is a random coincidence of stops, and
this occurs at a rate of approximately 4 Hz, so the probability losing data
is small.
For further details, contact Rich Baker or Doug Michael.