Choose timezone
Your profile timezone:
Status update
→
Europe/Zurich
6/R-018 (CERN)
Celebration Tobias got a staff post!
#################################
Michael , K-mod updates.
Q1A powering would solve the beta* control problem.
Beta at BPM from Kmod. Trying to improve the error bar at the BPMs beta in kmod.
#############################
Tobias, automatic coupling correction plans and OP
One issue is the binary conversion. This can be put in python merged with svd_clean. Lukas has this code!
They also want not to use afs. Should we go via a LSA jar?
Jaime says to include the jar in the beta-beating project.
The GUI mostly uses python so this should be ok.
gRPC is a tool to call python functions from Java. Could be useful for us.
Full response only for coupling ready.
ATS variable needed -> very important for commissioning.
How to speed-up? 10k turns, less lines from sussix? 100? We need sussix lines as input for drive.
Only calculate coupling in GetLLM?
1st then is the conversion+svd_clean and then will see.
Which python? Tobias will discuss with the python expert.
#################################
Elena, application
ADT feature in the model.
Flat for RDT analysis in global settings. Add a only coupling analysis flag.
Limit in BPM now increased to 2200. Number of turns should be 50000.
JPARC and PETRA flags new models implemented in GETLLM and in the GUI.
What about a miscelanea or Joker model?
Multiturn gui also has svd clean settings.
Kmod, orbit extraction with online model.
ADT works! There is still the problem of one missing BPM.
##############################################
Segment-by-segment in IR4
Andy finds that taking the right natural tune things work. Codes work!
##############################################
Lukas, randomized SVD
scipy.sparse.linalg.svds works! It is twice faster but has some lack
of determinism. Another method implemented Lukas.
Lukas will provide svdclean package that can read binary and can be faster
with this algorithm. Facebook already did it...
Lukas will release his version.
Stay tuned for future lukas' ideas
###############################################
Ewen, non-linear plans
in current plan a non-linear correction shift is considered
States of the power supplies is greatly logged in timber with RUNNING and IDLE.
The plan looks great.
Problem, on-line model only extract from main beam process, subbeamprocess are ignored in on-line model.
#################################################################
MD notes:
-Tobias 1: coupling doros
-Ewen 4: 2 are written, 2 have the results.
-Felix: Separate MD note for RDTs and AC dip aperture
-Ana: 2.5km
Journals:
-Lukas: ESRF
-Tobias: commissioning, suppression of DQmin (before commissioning? 1 day)
-Ewen: 4 papers? b4 before commissioning
-Felix: AC dipole DA, done
Report:
-Michael: K-mod , q1 split
-HL-LHC report: soon
There are minutes attached to this event.
Show them.