1. Preliminary analysis of PS septum 31 failure August 8th, 2008
Bruno Balhan, Jan Borburgh, Simone Gilardoni,
Yannis Papaphilippou, Antoine Prost, Rende Steerenberg

2. Event chronology Friday 8/8/2008
12:30 hrs; Fault SEH31, wrong current, wrong voltage
Vacuum activity observed in sector
During CT, beam lost around septum, while septum draws maximum current (power supply in current limitation mode)
Intervention in PS 14:00 hrs: HV cable broken, replaced during the afternoon.
During cable replacement, vacuum level hasn’t changed significantly.

3. Vacuum activity

4. Nominal CNGS intensity test
Friday end afternoon after cable replacement: few extractions with nominal intensity yields unchanged septum behaviour: power supply goes in current limited mode during extraction.

5. Re-setting up
Friday night OP does a fresh setting up of the CT extraction with reduced intensity:
Excessive losses still observed in the region of SS31.

6. Beam based observations Friday night
The extraction of other beams to TT2 (AD & LHCPROBE) worked well, only CT extraction was concerned.
CT related equipments:
BSW31
SEH31
BFA 9 & 21
QKE16CT
H8H16 + H420 beam control (H8H16 on other users too)
BSW16 (used on other users too)
Losses occurred during the first few milliseconds of the slow bumps and the QKE rise.

7. Beam based diagnostics of Friday night (1)
BSW31 verification:
Inom reduced to ~ 60 %  Losses reduced, but remained too high.
Closure of the bump was verified with orbit measurements  Bump was closed as normal.
Scan of BSW31 amplitude using pencil beam:
Losses increased for a too low value of the BSW31 current  Reduced aperture for circulating beam.
Losses did not really decrease after having increased due to septum foil. (other side of the septum foil losses should reduce again)
Mean radial position at extraction:
Verified using the 200 MHz reference frequency and scan made with radial offset.  Optimum re-capture at nominal radial offset.  Orbit ok.
Normal orbit measurements were made, disabling the de-bunching.  Seems to be ok.

8. Beam based diagnostics of Friday night (2)
BFA 9 and 21 excluded:
Pulsing at maximum of bump was verified and losses occurred a few milliseconds earlier.  Ok.
QKE16CT was replaced by QKE16 (different magnets and power converter) No change, QKE16CT ok.
Survey of the status of all other equipments was made in order to check if something was pulsing when it should not.  Nothing found.
SEH31 Anode and Cathode position and angle scan was made
Fault was found in the angle control and control card was exchanged, but this was not the cause as we always adjust using the AQN value.
Moving anode outwards (smaller extraction gap) reduced circulating beam losses.  Aperture reduction in circulating beam part for nominal foil position.

9. Beam based diagnostics of Friday night (3)
H8H16 + H420 beam control worked normally.  Ok.
In addition the BSW16, which worked well on AD beam and LHCPROBE beam was verified and found to be correct.  Ok.
Conclusion:
something in the septum had to be wrong.

10. Remote displacement system verification
Saturday 9/8/2008:
The link between CCC and the Central building electronics were verified: angle electronics card replaced
Access into PS confirms correct operational functioning of the remote displacement systems

11. Septum exchange
Over the weekend the septum kept the operational tension (180 kV), without drawing any current (No beam)
All previous investigations lead to believe the septum is at fault and needs to be replaced
SEH31 exchanged on Monday (11/8/2008) for new septum.
Withdrawn septum appears in good shape, except for the diffuser that seems to have lost 5 wires

12. Diffusor

13. Conclusions
Many diagnostics made pointed to a reduced aperture in the circulating beam part of the SEH31.
Five broken wires of the Difusor do not explain the presumed aperture reduction.
Septum too activated to make detailed visual inspection in the near future.
Final validation/diagnostics can only be done with beam, foreseen for Tuesday afternoon.