Choose timezone
Your profile timezone:
LHC FBs - weekly meeting
→
Europe/Zurich
Skype (Skype)
Skype
Skype
Meeting Date: 07-Apr-20 09:30
Location: Skype
Link to Outlook Item: click here
Invitation Message
Participants
Diogo Miguel Louro Alves (Meeting Organizer) - present
Jorg Wenninger - present
Stephen Jackson - present
Andrea Calia (Accepted in Outlook) - present
Michi Hostettler - present
Leander Grech - present
Notes
- Actions from previous meetings:
- BI:
- DA + LG: sanity checks of twiss parameters - not started but we know what to do and how to do it and we only do it for the elements in the mapping files
- DA + LG: BPM logging Vs feedback functionality - not completed - would require an enormous amount of code duplication and would be extremely error prone - proposal is not to implement this although we'll keep the changes that permitted the eradication of the bootstrap file
- DA: PPM settings of RefQQP property in BFBLHCSettings - done
- DA: Contact Jakub Wozniak about metadata of BPM data logged from the LHC FB system - done - basically for the new system we will have to log the mappings
- OP:
- AC: Provide new BPM mapping with extra column - not done and no longer required
- BI:
- Updates from BI
- More on the logging Vs feedback functionality action:
- DA + LG: implementing the logging Vs feedback functionality will be a nightmare. It will be an enormous amount of work, it will involve a lot of code duplication in order to have the 2 functionalities as disentangled as possible and it will be extremely error-prone. Thorough dedicated testing would be required.
- SJ + JW: better to have this complexity at a higher level (in YASP). The aim should be to simplify the low-level (FESA) side.
- JW: two approaches to bootstrap YASP with different lists of BPMs could be by using filtering objects (preferred by JW and in principle doable as confirmed by AC) or by pre-fetching the mapping from the BFC system.
- This action was proposed in order to simplify all matrix calculation tests on the YASP/Java testing infrastructure side
- MH: how should we handle masks when doing these tests?
- JW: some tests should be done with all masks OFF. Some other tests should be done with some BPMs and COD masks ON but always making sure that the Java and the Fesa sides use exactly the same masks. Like this, we should get good coverage.
- MH: how should we handle masks when doing these tests?
- Regarding the future logging in NxCALS and backward compatibility with data logged in CALS
- ACTION - DA: setup meeting with Jakub Wozniak, JW and SJ.
- SJ: revisit swapped BPMs - ACTION - DA - dig out past history and actions taken at the time
- ACTION - DA: setup meeting with Jakub Wozniak, JW and SJ.
- DA + LG: implementing the logging Vs feedback functionality will be a nightmare. It will be an enormous amount of work, it will involve a lot of code duplication in order to have the 2 functionalities as disentangled as possible and it will be extremely error-prone. Thorough dedicated testing would be required.
- More on the logging Vs feedback functionality action:
- Updates from OP
- AC: started work on FB on Thursday and Friday together with David. Had a bit of a nightmare configuring especially David's environment. Will continue until Wed 08/03/2020.
- AC: internet is getting slower and slower … pain to run VNC
- AC: tried running the FBs on his local machine
- AC: David managed to do some new tests (maybe just 1) on the optics - attempt to upload optics with negative ID
- DA: mentioned a few tests that can already be done including testing matrices and ref orbit and optics changes while executing the function player
- AC: will work on these in due time. Java API to handle the function player needs to be developed.
- Replacement of BPMR.5L8.B2
- JW: update LSA and FBs at the same time. On the LSA side to get a clean situation will take a few months in order to get a new machine sequence. Another trick could be to rename this BPM in LSA.
- JW: for YASP this will not be a problem.
- DA: from the Fesa side there should be no impact other than loading a slightly different mapping with this BPM replaced by a warm one (different name and different twiss values)
- DA + JW: in terms of logging, to be seen with CALS and NxCALS people how this should be done also in order to maintain consistency with already logged data - to be dealt with in previous ACTION
- Planning
- AC: current OP sprint will last until Wednesday. Will work on BFC until May. First task is to adapt YASP and then continue working on the framework. New tests will evolve at the same time.
- DA + LG: will continue working on the implementation of the missing functionality.
- DA + AC: will assess the feasibility of a joint sprint targeting the test coverage in due time.
- AOB
- LG: What about the implementation of the PID gains as function player waveforms?
- JW: would be nice to have the gain waveforms interpolated automatically in BFCLHC
- DA: this is not a problem and we will implement it as such
- JW: proposes same PID gain for orbit and energy loop and an independent one for the radial loop
- DA: which gains exactly (i.e. which value-items)? To be thoroughly checked in code as at the moment there's different gain possibilities (user-defined, etc).
- JW: simplify gain logic!!!
- JW: would be nice to have the gain waveforms interpolated automatically in BFCLHC
- SJ: would be useful to have a joint code review of the Fesa side later in the year.
- SJ: now that the bootstrap file is gone how is the system configured at startup?
- DA: element names and array and matrix dimensions come from the mapping files
- SJ: is it possible to switch the FBs on before loading valid optics and therefore dump the beam (as happened in the past)?
- DA: yes, that's exactly what will happen at the moment as the active optics matrices will not have good values - ACTION - DA + LG: check possibility to prevent turning FB on in case valid optics haven't been loaded … this is only an issue when the function player is not being used.
- LG: What about the implementation of the PID gains as function player waveforms?
There are minutes attached to this event.
Show them.
