CERN Accelerating science

Prototyping in the SPS using pickups equipped with LHC electronics

Meeting held on Friday 1-03-2002 - Chaired by Rhodri Jones, reported by Mike Lamont

Present : Lars Jensen, Pierre Charrue, Marc Vanden Eynden, Rhodri Jones, Joerg Wenninger, Thijs Wijnands, Mike Lamont

Agenda :

Progress with the SPS prototype

1. Hardware

Rhodri gave a summary of the state of the hardware construction and installation. His slides here. Things are progressing well and the pickups, cables, WBTNs and fibres should be ready towards the middle of the year.

2. VME

Lars gave outline of the contents of the SPS Real-time Orbit �BPL50S� crate. His slide here. Most things in place.

3. Controls

As agreed in a previous, unreported, meeting between Lars, Thijs and Mike, controls should provide:

• Interface based on UDP for data transmission to the top end.
• A real-time top end machine for data acquisition
• Software to display in real-time the orbit measurements in the control room (in the first instance). Clear that architecture/facilities of top end machine need to be established. We haven't done this before - one of the aims of the exercise.

Front end section agreed to make available a suitable machine within the week.

Thijs concerned about using a non-performant rickety old network which will necessarily not reflect final solution. Not thought to be a major issue at this juncture given the required data exchange rates. Agreed to try using existing network and then open issue to question should there be a problem.

Debate about where the manpower was going to come from. Might be able to secure use of Serguei who has done something similar for LHC power converter prototype in String 2.

Comments after meeting

Quentin

Just a comment on the comment:

3) Thijs will provide a small library that allows sending and receiving UDP packets on the standard SPS network. We see no need to complicate our lives with Quentin's JET stuff at this point in time. What I did for UDP communications at JET wasn't for no reason. What functionality will you be adding to UDP? At the very least it would be good to know if you are losing packets and how long they take to send.

Thijs 7/3/2002

I have made a simple INET library which is in /user/wijnands/IP You can choose between UDP or TCP, I recommend UDP for this work. I suggest you try to send some data from BPL40S to QLSBA3 (LynxOS 3.1.0) using my version udp_receive at the receiving end.

Lars 6/3/2002

1) One thing that was mentioned was that the mandate of this project is still not clear. Could Mike please clarify this point ... I believe that he gave a presentation at an SLTC meeting or ??? If we want anything more than a simple text output on the receiving end, it'll take manpower (CMW, graphical display) which may not be available. [see parent web page - Mike]

2) It was decided to use the already existing PowerPC called qlsba3 which is on the SPS network, in the computer cage in BA3. It is already configured with LynxOS 3.1.0 and is up and running.

3) Thijs will provide a small library that allows sending and receiving UDP packets on the standard SPS network. We see no need to complicate our lives with Quentin's JET stuff at this point in time.

4) For timing purposes we will ask for two GPS receivers from Garry. There's already an antenna on the roof of BA3 but none in BA5. Thijs will ask Garry if it's possible to move the antenna from BA2 where it's nolonger used, to BA5. Thijs seems to think that a time resolution of 1 msec is good enough. If we can't have the GPS, we can therefor use the timestamps from the TG8 distributed from the SPS MTG.