Minutes of the second WorldFIP insourcing meeting
9 April 2009, room 864-1c02
Present: Erik Van der Bij, Julien Palluel, Raymond Brun, Quentin King, Bruno Puccio, Ben Todd, Gonzalo Fernández Peñacoba, Javier Serrano
- Erik did a presentation introducing
WP3. This work package's deliverable is the functional specification
of the uFIP replacement, i.e. it answers the question "what do we
need?". Javier made sure the premise on which the presentation is
based, i.e. that we do not need a perfect copy of the uFIP chip, is
agreed upon by all attendees. Once this is agreed, we need to come
up with the list of features users require. Erik also presented a
preliminary version of a
survey
he intends to conduct in the coming weeks. The results will be
presented in the next meeting. Comments during Erik's presentation:
- Quentin: validate CRC in CernFIP chip before sending frames to user logic vs. sending raw bits to user logic. This is something to discuss during functional specs gathering. There seems to be wide agreement that the former makes more sense, since it puts complexity in the CernFIP chip, which is a carefully studied development where this thing would be solved for every user. Delegating this to the user logic will result in many solutions with varying degrees of performance and robustness.
- Quentin: we should also discuss an 8-bit vs. a 16-bit interface.
- Quentin: bandwidth is currently wasted due to the FDM implementation of presence checks. We could try to propose an alternative solution.
- Erik: out of scope of this project, but users should check that they are properly handling uFIP hangs with appropriate power cycling techniques.
- Gonzalo: it would be good for designers to have access to a catalog of these techniques as currently used in different systems at CERN.
- Gonzalo and Quentin: try to see if we can put the digital part of the FielDrive inside the CernFIP so as to replace the FielDrive - MicroFIP combination by a standard RS-485 transceiver plus the CernFIP. Javier says this will need some effort to understand if the FielDrive is doing any analog tricks we're not aware of, but in principle it looks like a good idea, since this project is really concerned with the FielDrive - MicroFIP combination and not the MicroFIP alone.
- Erik: the fieldbus standards have been purchased by the CERN library.
- Julien presented the plan for testing
BE-CO WorldFIP hardware in the CNGS facility. The fact that
repeaters are being tested independently will give us for the first
time a way to distinguish between the radiation hardness of the uFIP
and that of the FielDrive. Comments:
- Erik: look for opportunities to replace old Cu/Cu repeaters in not-so-hot areas by the new version (apparently less rad-hard because of new FielDrive) so as to gather a big enough stock of old repeaters we can then use as replacements in hot areas.
- Javier: we need to check whether CNGS will be operating properly in week 18, when the tests are scheduled, after the earthquakes in the Abruzzo region.
- Next meeting in about 4 weeks with a report on Erik's findings for WP3 and also some words on Raymond's and Erik's trip to Paris to gather some more documentation from Alstom.
J.Serrano, Apr 2009