Log entry time 23:53:40 on April 20, 2004

Entry number 82723

Run Coordinator's Report for 19 April Swing to 20 April Swing

Summary

Significant News

• The half gain HV file was not downloaded correctly, causing the setting to be about 100V higher for some channels. This caused those detectors to have higher pion rates, but left the proton (elastic and inelastic) unchanged. After turning the HV OFF and recycling the power on the CAEN the download of the file worked and the symmetry in the octants was restored. See Julie's and Benoit's log entries.
• Took production data during the night. Beam quality was not all that good (lots of beam trips, Kaz's feedback system had trouble). There was also a CFD unplugged for half the night (see log entry number 82672). All in all a rather disappointing night.
• Beam studies from 09:00am to 12:00noon. Recovery from beam studies turns out to be VERY painful. Parity quality is very poor, Kaz had a lot of trouble with the feedback due to large PZTX induced charge asymmetry and slow x-position response to PZTX and PZTY changes. There seems to be some problem with the beam transport in the Hall. See the log entries by Kaz and the ELOG for more details.
• During the beam study period it is discovered that the raster is on the pivot focus instead of the G0 focus. We do not know how long this was the case but RC will try to find out if there is a log of this information somewhere. See Greg Smith's log entry for more info.
• Still no beam delivery after Kaz is finished with his IA/PZT scans, due to Linac RF problems. The beam spot is more like a beam stripe (see ELOG entry). The problem is finally resolved ( ELOG entry) and we get beam at 22:45. We start to take production data.
• Finally on a good note: IOC SE20 seems to be more stable than before. The swap on Monday seems to have helped.

Short term Plans

• Continue with production data taking on LH2 until beam studies period tomorrow morning 09:00am when beam will taken away for beam studies and Hall access.
• Work during the access on Wednesday morning:
  • Cryo line of the magnet (Damon Spayde)
  • Installing electronics for cavity monitors (John Musson)
  • Swapping electronics for Oct1&3 with Oct5&7 (Paul King).
• Beam energy measurement Wednesday afternoon at 3pm (if beam is restored in time). Mark Jones will be present for this.
• Moller/IHWP insertion/Moller every 3 three days.
• Continue with LH2 production running after downtime and energy measurement.

Longer term Plans

• Moller/IHWP insertion/Moller every three days.
• NO beam studies on Thursday. Beam studies on Friday from 09:00am to 10:00am.
• 24hrs of aluminum frame target running. Discussion ongoing.
• Deuterium target(?)
• Long 10uA running for deadtime studies(?). This is still under discussion, also because Brian Quinn and Angela Biselli proposed a different way to extract the "deadtime" of sub-threshold particles.

Concerns

• What raster focus were we running with in the last few weeks?
• After replacing the IOC SE20 controller board there are still a few IOC SE20 crashes, but they have gotten a lot less frequent (about once a shift or once every 2 shifts). The IOC crashes in way that makes it hard to detect: The EPICS values freeze at a non-zero value. If this happens the IOC has to be rebooted. Paul King has written a script that gives an audible alarm if the G0 BPM flatline.
  When the IOC crashes the run should be immediately stopped and a note put in the end-of-run comments and in the run summary that states that the IOC crashed at the end of the run.
• Temperature in the electronics cage is rising during the afternoon hours above alarm levels. The big fan in the cage should be started to increase airflow to the FR electronics rack. See Paul King's log entries Monday and Tuesday.
• Spikes in French electronics, possibly caused by a network problem. Paul King claims that the buffer of ROC3 may not be big enough (currently ~2sec) and could be increased up to ~3sec, but then we would have to boot ROC3 every time we change the configuration. Shift crew should also clearly mark runs with spikes.
• NA LTD errors still occurring at a small rate. Brian has generated a "HowTo" document for trouble shooting.