Re: [BLAST_SHIFTS] Time stamp on physics event fragments

From: Karen Dow (kdow@mit.edu)
Date: Thu Dec 12 2002 - 10:13:27 EST


        The new readout lists are

/home/blast/blast/commis/coda/crl/commis_l_pri_time.crl
/home/blast/blast/commis/coda/crl/commis_r_pri_time.crl

The object code is on node blast in /usr/local/blast/daq/obj. If you
need to go back to the readout list without time stamp, follow the
README in blast:/usr/local/blast/daq to remove the symbolic link and
link to the usual object code. Then you'll have to use RunControl
Abort, Download to pick up the changed readout list.

        The boot scripts for the ROCs that don't have the real-time-clock are

blast:/usr/local/blast/vxWorks/vxWorks-startup-blrocf2-notime
blast:/usr/local/blast/vxWorks/vxWorks-startup-blrocf3-notime

I don't think there is any reason to go back to the old boot script, but
if you have to, copy the -notime version to vxWorks-startup-blrocf2 (or
3) and reboot the FASTBUS ROCs. Then RunControl Download, of course.

        I looked at the readout busy time (Level 1 Accept from the Trigger
Supervisor) for the time stamp case (running some kind of 30Hz cosmic
trigger), and it was about 750 usec. That's about what it was before
(without the time stamp), I recall. It may be that high physics rates
will cause the clock to be less accurate, we'll have to see. Each ROC
checks its time against dblast, and corrects as necessary, every 60
seconds. The correction is printed out to the ROC serial port (or you
can see it by telnetting to the ROC); I have to look more to be sure of
the units on the correction but I think it's seconds.

                                        Karen



This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:29 EST