Transcript Review_4-5_rd-Reiner - Indico

AT-MEI-PE, R. Denz, CERN, CH-1211 Geneva 23
Hardware Commissioning of Sector 4/5
Quench Protection System
R. Denz AT-MEI-PE
1
Prologue
 Compared to the 7/8 exercise significantly less problems  time to
introduce the QPS field team (the blue button on the war map …)
Mathieu Favre: Installation, QPS local cabling and internal field-busses
Vincent Froidbise: Digital quench detection systems
AT-MEI-PE, R. Denz, CERN, CH-1211 Geneva 23
Joaquim Mourao: Main magnet protection, quench heater power supplies
Grzegorz Seweryn: Installation, equipment preparation
2
600 A circuits
 Optimization of detector firmware lead to significant progress
– 0.1 As-2 and 5 As-1 reached with RSD2.A45B2 (12 x MS, 0.4 H)
– Inductance tables turned out to be very precise (Grazie Walter!)
– Rather limited time for test & developments
AT-MEI-PE, R. Denz, CERN, CH-1211 Geneva 23
 Voltage “spikes” during 0 V crossing of 600 A power converter (see also talk
given by Yves)
– Triggered QPS very often
• PC distortion:
U_distortion x t_distortion = 2 V x 1 ms
• QPS threshold: U_threshold x t_discrimination = 0.1 V x 20 ms
– Special non-linear digital filter developed, implemented and successfully
tested in lab
 No information yet on requirements for fast feedback
– Fast feedback should have been tested in 4/5 – it will be a must for 5/6
3
Firing heaters without current in the magnet circuit
 Effect observed only in main dipoles after high current ramp
– Not a spurious trigger or caused by EMC – recorded signals are
consistent and real
– Probably triggered by local heating of magnet causing a break-down of
the persistent current
AT-MEI-PE, R. Denz, CERN, CH-1211 Geneva 23
• One case observed after accidental firing of quench heaters
– Order of magnitude corresponds to theoretical predictions ( N.
Schwerg, S. Russenschuck)
 Main magnet quench detection systems have no information about current
level
– System reacts correctly by firing heaters
– No safe work-around
4
Feedback from QPS teams
 Up to about 30% idling time noticed – even when working with “only” two
fronts
– Teams waiting for signatures, other experts, end of the RAT meeting …
– Most critical times: 7h00 – 9h00 and 16h00 – 18h00
– Equipment / system downtime to be given more precisely
AT-MEI-PE, R. Denz, CERN, CH-1211 Geneva 23
 Limited flexibility of the sequencer application
– More people should be able to change circuit parameters
– Successfully performed hardware tests should not be repeated because
of software issues
 Tests requiring presence of field teams in the tunnel shall be executed with
high priority
5
Conclusions
 Hardware commissioning or 4/5 fully successful for QPS and EE
– All circuits efficiently protected – no damage
– All systems reached design parameters (or more …)
– Number of accesses to LHC tunnel significantly reduced
AT-MEI-PE, R. Denz, CERN, CH-1211 Geneva 23
– More trained QPS (hardware, firmware) experts available
– Some very experienced QPS experts (including two team leaders) will
leave during 1st half of 2008
• Four new experts for CCC operation required (4 front baseline)
 Preparation of QPS for LHC operation has started
– QPS must become almost transparent to LHC operation
– Macros, automatic re-arming after a PM event …
 See you soon in sector 5/6, 7/8, 8/1 …
6