Shift summary for owl shift 11.18.99

Jim Graham (jagraham@hep.uchicago.edu)
Thu, 18 Nov 1999 06:54:33 -0600

Shift Summary for Owl shift, 11.18.99

On Shift: JAG and JS

Hours of beam delivered at all intensities: 6:20
Hours of beam delivered at nominal intensity: 5:40
Hours of E799 data (special runs not included): 6:00
Hours of special runs (laser scans, muon runs...): 0.20

Accounting of lost time:

========================================================================
Runs ended this shift: 14997-15000

Run 14997 - Muon run, ended on DDD vetoes
14998 - 200 k muon run
14999 - Beam 799, 1.2 M events, ended to adjust gain on CsI xtal 2893
15000 - Beam 799, 1.6 M events, ended to adjust gain on CsI xtal 2893

========================================================================
Accesses during shift: None

========================================================================
Overview of shift:
Smooth E799 data taking, beam oscillates between 4 - 8.5 E 12. Beam
has wide horizontal seed position; MCR can't seem to fix. Crystal 2893
(which had HV divider replaced on prev shift) had to have the gain q-to-E
constant adjusted twice, for cumulative correction of 0.74

========================================================================
Description of Shift:

run spill time comment
------------------------------------------------------------------------
14997 12:39 beam returns
1 triggers hang on all four DDD vetos. drt_fresh_start

14998 1 again, DDD4 veto. This time, page Ed. He says to
clear busy logic on ktrig window. This works for a
few spills, enough to get 200k of muons. This is
probably due to high rates of muons.

14999 1 1:12 start beam 799 run
1:35 after multiple Kfrends and pipeline inits, run is
finally going smoothly. No DDD veto problems.
1:59 Check CsI gains on xtals 2871 and 2893. 2871 looks
fine; <E/p> = 0.9895, based on 3 electron clusters
with this as the seed block. Xtal 2893 looks low;
E/p = 0.81, (based on 1 electron); however, the 8
neighbouring blocks all have E/p 1-5% lower than
their cumulative average, probably pulled by xtal
2893. Page Ed. He decides to adjust the q-to-e
constants for this channel by 0.8. Do this, and
stop and start run to pick up new constant from dbase.

15000 1 2:20 Start run
2:48 Clean up onsplit disk space by deleting the following
caches marked for deletion: UPF038, UPH025, UPK019,
UPLA07, UPM030.
3:14 Find CsI xtal 2893 has E/p =0.928, so adjust by
this factor again in q/E constants. Stop run

3:00 The C02 dewar is at or below 100; it may need to
be checked by a DC expert.

15001 1 3:20 start run
3:38 call MCR about bad horizontal beam structure. We
have been getting SDAQ alarms all night about scaler
ratios just out of limits, which I believe is due to
this problem.
3:40 Find channel 2893 at E/p=0.9873 -- not messing with
that.
4:57 MCR returns phone call; they can't find the problem
with the horizontal beam width. Will continue to
look.
5:39 Get a few Elliott HCC ineff audio alarms. No
obvious HCC bit-pattern problems, but 3 bits have
cumulative ineff > 1% (near small/large boundary)
Reload Et.
5:41 Notice that V1 East source not present in latch
distributions. Also EWUD logic not 100% efficient
due to V1 east. Since EWUD isn't used in any 799
trigger, this is not an immediate problem, but may
be one in the future.

========================================================================

Follow up Notes on Previous Shift
CsI crystal 2871 and 2893 were monitored with electrons to see if the
laser gain change was followed by electrons. 2871 looked fine, (indicating
that only the fiber had moved), while 2893 looked low in E/p. It took
two runs to adjust it to its current value (based on 1 electron) of
E/p = 0.9873.

========================================================================
Items to Follow up on Next Shift
Monitor channel 2893 in csigain.
long term: fix V1 East source/trigger ?