Minutes of the 2^nd Timing Working Group Meeting (TimWG)
Held on Friday the 30^th April 1999

  Present:

Excused:

Rudiger SCHMIDT, Philippe BAUDRENGHIEN.

Agenda

• Mandate, all
• Web page, Etienne
• SL/PO requirements for LHC, John
• SL/BT requirements for LHC, Etienne
• Next speakers
• Next meeting

Mandate

A general review of the proposed mandate has been made. Two modifications have been decided:

• "The summary of the WG should be documented as a functional specification, possibly for the end of September 1999" has been replaced with "The summary of the WG should be documented as a functional specification, for the end of September 1999".
• "The WG should formulate recommendations in order to avoid divergence between solutions for the different systems" have been replaced with "The WG should formulate recommendations in order to avoid unnecessary divergence between solutions for the different systems".

Web Page

Etienne CARLIER presented the different Web pages created for the working group and the two different mailing lists. It has been decided to activate the link in the LHC Working Groups page.

SL/PO timing requirements

John PETT explained the different timing requirements for the SL/PO group (See slides attached) and  the status of the envisaged solution. He presented a timing distribution infrastructure (See slides attached) based on a timing gateway (PowerPC, TG8, GPS) connected on one side to an Ethernet TCP/IP network for remote control and monitoring, and on the other side to an industrial WorldFIP field bus for timing events distribution. Timing events may be received through a TG8 card or, by the control network and distributed to the different clients by broadcasting messages on the WorldFIP field bus. A typical event is a command to activate the requested action at W hours, X minutes, Y seconds, Z milliseconds, and is transmitted approximately 100 ms before the required time. Repeat message mechanism and client feedback to the master will be implemented in order to guaranty and supervise the correct event distribution. Approximately 10 us jitter on the event distribution over the field bus is hoped for. The main advantage of this approach is the cost reduction of the power converter timing distribution (less hardware implemented and less cabling work) and a global simplification of  the complete setup. Michel JONKER suggested that this approach can be an alternative solution for the "1 ms" timing distribution. Jean-Jacques SAVIOZ said that SL/BI group is actually evaluating the possibility to use the same approach. A test setup of this solution will be turned on during the second semester of this year. John PETT concluded that this proposal is mainly based on the SL/PC requirements and not the definitive solution.

SL/BT timing requirements

Etienne CARLIER presented a general overview of the SL/BT timing requirements (See slides attached). The more important points can be summarized as follows:

• Actual functionalities of MTG and pre-pulse distribution fulfill basic requirements for the SL/BT group,
• Fast timing pre-pulses are necessary for injection kickers and must be synchronized with injected and circulating beams,
• Few amount of slow timing events are needed,
• An important parameter is the LHC revolution frequency (extraction kickers and dynamic aperture measurement kickers),
• Need for a redundant fail save and/or fault tolerant timing distribution system (mainly for the LHC extraction kickers),
• Need for a deterministic timing distribution (which event, where and when) for the fast and slow timings but also for the revolution frequency,
• Need for a synchronization mechanism with other equipment:
• slow synchronization for continuous data acquisition and logging during normal operation (msec., sec.),
• fast synchronization for post-mortem analysis after equipment failure or beam dumping request (usec.).

Next meeting

Next Meetings will be held  on Wednesday 12 May 1999 at 10:00 AM in 865 1B03.
 
Agenda

1. Approval of previous minutes
2. CMS test beam timing requirements. Emmanuael Tsesmelis
3. SPS Master Timing Generator requirements. Michel Jonker
4. Next speakers
5. Next meeting
6. AOB