Minutes of the 7^th Timing Working Group Meeting (TimWG)
Held on Friday the 9^th July 1999

  Present:

Excused:

            Philippe BAUDRENGHIEN

Agenda

• Approval of previous minutes
• Timing transmission over the WorldFIP bus. Quentin KING (SL/PO) & Raymond BRUN (SL/LRF)
• Next speakers & next meeting

Approval of previous minutes

There was no comments on the previous minutes.

Timing transmission over the WorldFIP bus

Quentin KING presented the possibility to use the WorldFIP fieldbus to synchronize the clocks on all LHC power converter controllers to better than 1ms compared to an absolute standard, to maintain this synchronization with a digital phase lock loop and to distribute "events" based on absolute time (see slide attached).

He explained the system architecture based on a three layers system (worksations, gateways and power converter controllers) as in LEP, justified the selection of the WorldFIP fieldbus mainly for its deterministic traffic capabilities, evaluated the architecture of the middle level gateway with and without a TG8 card and explained in details the functionality of the digital PLL which guarantee a jitter better than 20us for the complete synchronization chain. He specified that an event transmission requires a minimum warning time of 100 ms, that an event can be emitted each 10 ms and that an specific communication protocol guarantee the event generation and reception. Events can also be sent to specific groups of power converters or to an individual power converter.

The general philosophy of timing transmission over WorldFIP proposed by Quentin KING can be summarised as: "send an event when you want and execute it when you like".

Raymond BRUN presented a solution based also on WorldFIP fieldbus to distribute with a very high accuracy the time to different equipment. This solution is based on the WorldFIP fieldbus deterministic traffic capabilities and on an analog phase lock loop which permits to transmit the hour with a precision better than 10us. With this solution machine events can be dated rapidily with a very high precision.

AOB

During the following discussion,  technical questions to be discussed in the next meetings have be raised:

• Machine state distribution?
• Broadcast requirements?
• Timing distribution requirements (determinism, reliability...)?
• Time format?
• Timing diagnostic (tools, facility...)?
• Sofware timing?
• Computer synchronization?

Next meeting

Next Meeting will be held on Friday 20 August 1999 at 10:00 AM in 864 1C01.

Agenda

1. Approval of previous minute
2. Review of the different timing requirements and locations.
3. Next speakers & next meeting
4. AOB