Minutes of the 14^th Timing Working Group Meeting (TimWG)
Held on Friday the 24^th October 2000

Present:

Agenda

• Approval of previous minutes
• SPS timing in LHC, Gary BEETHAM
• IRIG-B norms, Christoph KNAUPP
• Time of day standards, Gary BEETHAM
• List of slow timing requirements, Mike LAMONT
• AOB

Approval of previous minutes

There were no corrections on the previous minutes.

SPS timing in LHC

Gary BEETHAM reviewed the status of  the different SPS timings (slow and fast) to be distributed in the LHC (see slide attached). After some discussion, the following propositions have been accepted:

1. SPS fast extraction prepulses will be distributed to LHC point 2, 4 and 8 (see slide attached).
2. SPS revolution frequency will not be distributed to LHC.
3. SPS MTG timing events useful for LHC operation (i.e SPS RF timing events or SPS extraction warning events...) will be available in LHC through the LHC slow timing distribution system.

Thomas BOHL reminded the wish of the SL/RF group to have in the LHC one single distribution system for both SPS and LHC timing events. During filling the LHC RF needs timings similar to the ones sent on the SPS MTG in order to control RF equipment which are also used during the LHC ramp (cavity voltage or the transverse dampe). For this reason, SL/RF would like to avoid two different timing distribution systems (SPS during filling and LHC after start ramp) driving the same equipment. Philippe BAUDRENGHIEN added also the necessity for the SL/RF group to keep in the LHC slow timing distribution system the functionality actually available in the SPS TG3 (enable/disable event, delay event...). Gary BEETHAM confirmed that this kind of  functionality will be available in the LHC slow timing.

IRIG-B norms

Christoph KNAUPP presented the different formats possible for the serial distribution of the time of day (see slides attached). Three possibilities exist:

• IRIG Serial Time Code Format - IRIG Standard 200-98
• Norme Francaise NF S87-500
• Homebrew Version based on IRIG Standard.

Each format has a different bit pattern, but the first part of the time code frame for all the format contains always the time-of-year in binary coded decimal (BCD) notation   (days, hours, minutes, second). One major difference between the two first possibilities is the distribution of the year information. Year information is available in the Norme Francaise code frame but missing IRIG-B code format. In order to maintain the compatibly with the two industrial norms, Gary BEETHAM suggested to avoid the development a specific CERN code frame (Homebrew Version based on IRIG Standard) and to used only the BCD value of the  time-of-year. Distribution of the time-of-year up to the PLC will be discussed later once the requirements from the post-mortem system will be identified.

Time of day standards

Gary BEETHAM presented the different time of day standards (see slides attached) and explained briefly the historic of these different standards. He outlined the problem of time difference between the standards (mainly TAI, GPS and UTC) due to the introduction in 1972 of leap second mechanism in order to compensate earth deceleration.  In order to avoid different time of day reference around the machine, Gary BEETHAM recommended that the CERN LHC Time of Day standard will be referenced to   Coordinated Universal Time (UTC).  Julian LEWIS expressed the compatibility of this proposition with the UNIX operating system time reference which are also linked with UTC. The working group members have agreed on this proposition.

List of slow timing requirements

Mike LAMONT reviewed the different categories of slow timing events necessary for the LHC operation:  ramp events, measurement events, synchrone and asynchrone events, injection events, RF events, post-mortem events (see slides attached). Taking into account these categories, he identified  the principal issues that have still to be clarified before a more in-depth analysis of the different possible technical solutions.  He analyzed also briefly the advantages / dis-advantages of the different timing distribution system alternatives (classical timing, BST based, absolute time and real time) presented by Michel JONKER in a previous meeting (see minutes). Gary BEETHAM explained that each of these alternatives will be discussed in more details during the next meetings. He reminded the clear need for a uniform distribution mechanism of the commonly used machine data (machine mode, beam current, energy...) and raised up the question if these data can be distributed through the timing system and outlined the questions of the timing system reliability. Regarding the reliability of the timing system, Robin LAUCKNER explained that, for him, equipment have not to rely on a 100% availability of timing event and that interlocks have to be foreseen in order to protect equipment/machine against missing events.The necessity of checking event distribution up to the user (handshaking) has to be discussed with the LHC interlock manager. 

AOB

Next meeting

Next Meeting will be held on Wednesday 1st November 2000 at 14:30 in 864 1C01.

Agenda

1) TTC costs and availability, Bruce TAYLOR
2) Fast and slow timing over TTC, Jean-Jacques SAVIOZ
3) No timing network, Julian LEWIS
4) Implication of timing distribution over common purpose network, Michel JONKER
5) AOB