Log entry time 13:47:38 on March 02, 2004

Entry number 78432

RC Summary

Monday, 1 Mar. 2004 day through Tuesday, 2 Mar. 2004 owl

• Data taking on LH2 target with 40 ľA and with 6 mm hole target.
• IHWP is OUT. Next change on Tuesday 2 March.
• LH2 target parity quality
  • 15.6 hours for this 24 hr period (runs 20372-20400)
  • Overall efficiency: 38.8% (9.3/24) (pretty poor)
  • Beam (no trip) efficiency: 79% (12.3/15.6)
  • Parity efficiency: 76% (9.3/12.3)
  • Total parity hours: 182.3 (runs 19815-20400)
  • Overall IN/OUT charge ratio: 0.89

Achievements and events

• Production running at 40 ľA and running with 6 mm halo target.
• 120 Hz runs taken during
  • Monday day shift: run #20383 analyzed
  • Tuesday owl shift: run #20395 analyzed
• Coil modulation runs taken during
  • Monday day shift: run #20379
  • Monday swing shift: run #20385
  • Tuesday owl shift: run #20396
• G0Integrity results recorded during the
  • Monday day shift
  • Monday swing shift
  • Tuesday owl shift
• Convergence of beam parameter asymmetries checked
  • Monday day shift
  • Monday swing shift
  • Tuesday owl shift
• The electronics cage temperatures made a serious upward swing near the end of Monday day shift. See also Louis' log entry,

Plans

• The bulk of the week should be devoted to LH2 production running.
• We will stop at roughly 9:00 AM every day for 1/2 hour to perform Chow's fOps test plan until further notice.
• A polarization sequence (Moller measurement, followed by IHWP reversal (to OUT), followed by another Moller measurement, and then a RIP scan) will take place, as normally scheduled, on Tuesday March 2nd.
• We would like to do one energy modulation (in addition to coil modulation) run per day. Permission has been obtained from the Hall B Run Coordinator, but the shift leader should inform the Hall B shift crew before begining the modulation. (We don't think that they will be able to see anything.)
• We may do some test to characterize the BPM misbehaviors we're been chasing. A tentative plan for a study is available, but nothing is scheduled pending more input from experts and other interested parties.

Concerns and Their Resolution

• Three tests of coil modulation with the halo target in have been carried out in preparation for "auto-coil-modulation" at the beginning of each run. The earlier ones showed large increases of halo rate during the scan which indicates that the beam is either not centered or has a large halo, but on the most recent one, the increase was limited to a factor of ~2. For now we will continue doing coil modulation with the halo target out. Given the large difference in halo rates between when the 6 mm hole target is IN and when it is OUT, we can conclude that there is a beam halo. Further more, it is larger than it was in early February by a factor of 2 or 3 (see Log Book Link and Log Book Link). We noticed a marked improvement in the halo rates when the beam focus was changed during the quad scan carried out today. All of this suggests that we are actually seeing halo and that it should be possible to improve it, though by the rough calibration, we are well below our spec. We need to be absolutely sure that 1) the halo is not affecting our data and 2) the halo target is not affecting our data.
• We seem to be seeing more DAQ failures of late (see Log Book Link, Log Book Link, and Log Book Link for example): something like once per shift. The procedure for getting things going again usually involves reloading all of the various ROCs and processes, a procedure that can require 15-20 min (perhaps longer for inexperienced shift leaders). Is this still a rate related issue (though our rates are not so high now)? Is there any way to diagnose these crashes and work toward eliminating them?
• Spikes in the French timing spectra appear from time to time. Louis is here and will be investigating (Log Book Link). The spikes seem to have grown scarcer of late, but no explanation yet.
• More North American electronics errors have occurred (Log Book Link). One of the offending LTDs was replaced. It had a lose connector, possibly due to the currently practiced method of eliminating errors (cable wiggling). The error rate seems to be somewhat improved now.
  
• The issue of variation of BPM response with current has been investigated by looking at the replay and the EPICS archive (Log Book Link). Indeed, for the Y coordinate of the girder BPMs there appears to be a problem when using the 20 uA pulsed beam. We need to find a way to run the harps at higher current to find the proper BPM readings when we are doing real data collection. It is also possible that the BPMs behave differently with and without the raster. A test of this idea will await a convenient point in the schedule. Steve WIllimason has been contacted by Richard Dixon, who had some useful information. Testes may be forthcoming.
• A couple of runs (20246 and 20291) showed a large reduction in rates on both French and North American detectors. Paul traced this problem to a brief shift in the Musson box timing shift by about 0.5 ns. It is not clear what caused the change in timing. There has been a slow trend toward lower elastic proton rates also seen on both detector packages, perhaps this can be explained by Musson box shifs at a low level. The data that we have acquired so far should be checked to determine how often Musson box shifts happen. We also need a way to determine in real time that there has been a shift (Realtime monitor plot? EPICS alarm on the control voltage? StripTool plot of the control voltage?)
  
• A study of the effects of the 70-sec trip delay on asymmetries (Log Book Link) indicates that we should be able to shorten this delay to something like 20 sec (500 mps). More checking needs to be done. This would be good for us (more data) and good for our relations with MCC.
• We should spend some time during the 4-day maintenance period to try to find a way to turn the magnet power supply off at full power without causing a fast dump and without compromising our quench protection.
• IOCSE20 seems to be needing more trips. A older, slower (but hopefully more robust) replacement is available. Time to install ~2 hours. No guarantee of that it will help. We probably want to have it put in during the maintenance period, but don;t want to use beam time unless the problem becomes worse.

Raw data: List of shift summaries

• Monday day shift: Summary and ABU report
• Monday swing shift: Summary and ABU report
• Tuesday owl shift: Summary and ABU report