Mar 23/23 PionLT/KaonLT Analysis Meeting Notes ---------------------------------------------- (Notes by GH) Today: PionLT will be discussed first Please remember to post your slides at: https://redmine.jlab.org/projects/kltexp/wiki/Kaon_LT_Meetings Present ------- Regina - Garth Huber, Ali Usman, Alicia Postuma, Muhammad Junaid, Nathan Heinrich, Vijay Kumar ODU - Carlos Yero JLab - Dave Gaskell CUA - Richard Trotta, Tanja Horn Ohio - Julie Roche Carlos Discussion ----------------- Carlos has agreed to attend and present some of his Lumi scaler analysis results, so he goes first Carlos discussed scaler analysis results with Richard, and also redid some KaonLT lumi scaler analysis - plots of T1(SHMS-3/4)/mC, Runs 5157,56,55,54 positive polarity runs - sees steep rise with current to 55uA - plots of T3(HMS-ELREAL)/mC, same runs (electron polarity) - sees flat dependence with current - also plots T1 vs scaler rate, T3 vs rate - expects an increase with rate because of increasing randoms and scalers of course can't eliminate them Carlos shares his findings from CaFe analysis: - motivation is to understand why there is a drop in yield (counts/mC) with beam current observed in CaFe results. The yield has been corrected for HMS/SHMS track eff, total LiveTime His slides on relative yields rate dependence are available at: https://hallcweb.jlab.org/DocDB/0012/001209/001/rate_dependence_slides_cyero_Mar09_2023.pdf Slides on Event Type Definitions at: https://hallcweb.jlab.org/DocDB/0012/001208/001/event_type_definitions_COIN_DAQ_cyero_Mar16_2023.pdf The slides on event type definitions might be helpful when trying to understand how to properly select which event types based on whether singles or coincidence was taken, and how to extract the singles and coincidence separately. This is sort of complementary to Jacob's EDTM studies. It might help in selecting the right event types when calculating EDTM live times for singles, in which case you would need to apply the proper event type cuts. - in CaFe trigger config, T5 (coin) is formed by overlap of SHMS (T2) and HMS (T3) - 1st hit in look-back window gets selected and is used by the analyzer to form a coincidence time (i.e. T2-T3) - it is needed to set a cut on the raw TDC spectrum for the relevant channels (T2, T3), otherwise an accidental at arrives first in the look-back window will "block" the real coincidence - generic parameter file: PARAM/TRIG/tcoin.param - coincidence time (corrected for path length) without this cut looks drastically worse than one with properly set cuts - also, the number of events with good CoinTime is higher with good cuts, as fewer coincidences are blocked by accidentals - the relative yield from Be9 target versus current is MUCH different after the cut is optimized - BEFORE: ~7% decrease in yield is observed to 60uA - AFTER CUT OPTIMIZATION: the relative yield is nearly flat with current, at most a ~1%(+/-1%) effect - Conclusion: it is EXTREMELY IMPORTANT to verify that these cuts are set correctly (see his example TDC spectra) as otherwise there will be a large rate dependence and it will be hard to diagnose what is wrong! Richard Continues this Discussion --------------------------------- - Nscaler = sum(trig scaler) - EDTM_scaler - Yield_scaler = Nscaler/Q - runs 5154-58, 5298, all same setting Compare Richard & Carlos results - two main differences 1) when using same current cuts as Carlos, linear dep is restored, same as Carlos 2) HMS: RT gets linear increase ~5%, while CY sees flat - also RT gets rates up to ~100kHz, while CY has rates to 60kHz FOR SAME RUNS - CT data are middle 4 of RT's 6 runs ** RT, CY will meet to compare exact HMS scaler values on Friday, and see where the rate discrepancy is coming from. Ali will join online ** Nathan will check PionLT data to see that trigger cuts that Carlos described are set correctly ** it is important for RT, VK can confirm that these cuts are set correctly on KaonLT and Summer-2019 data - check both TDC and yield vs current for a solid target - CY: also be sure that EDTM-LT cal uses EvtType >= 4, to get all COINS when counting EDTM-accepted Richard Updates --------------- 10.6GeV carbon lumi scan analysis - Lumi #3, 5351-58 (HMS only) - plots of Yscaler, Ynotrack, Ytrack - get a consistent HMS current trend, ~2% increase to 65uA - DG: most reliable is Carbon Track, which is cleanest - worries about pi- contamination at high rate in scaler-no-track - Carbon Track seems flat within errors (Runs 5154-5298) - TH: this looks good, please move to CryoTarget study - GH: it would be good to check 8 or 6 GeV (whichever is more reliable) so that we know things are consistent New script that lists the cuts applied to a given run in python anaysis - UTIL_KAONLT/bin/cut-check - used to check what cuts are used to trim the tree created by python analysis - hopefully this will be helpful in reducing errors, as otherwise one has to go through a lot of files to list all of the cuts being applied, which sounds like a recipe for error RT, AP, AU meeting tomorrow to discuss the large right SHMS setting rate discrepancy between the two analyses reported last week Nathan Updates -------------- Aerogel Calibs - last report: PMT4+ calib had a slope vs run# - broke up into 3 sections - 2021-22 run now 2 parts, 1st part is most of 1st beam energy - summer 2022 is 3rd part - takes simple mean for each period, not an error-wt average, since errors seem unreliable - excluded some runs where NPE histo fits bad - put new params in hcana DBASE - replayed data w/new calib has better pion-separation from zeros in Aerogel compared to online replay Still waiting on Junaid for RefTime cut check GHP presentation - hoping Tanja has scaled cross section values from Scaling Paper at x=0.311 otherwise he will have to read them off the plot, which is less accurate - will circulate draft slides next week Junaid Updates -------------- DC Calibs - found some errors in PionLT.database file and kinematics file - some missing run numbers and typos - SHMS DC look ok - HMS DC calibs off, still needs work - getting high #events at zero bin entry - DG: what kine of runs are these? Production - part of the trick is to apply electron cuts to electron arm, otherwise get a lot of junk that distort the DC spectra - calib code is awkward, as you have to "hard wire" the PID cuts - MJ still checking things, HMS plane residuals improving Ali Updates ----------- First look at pion-PID for high Q2 data - shows plots where progressively apply cuts - get a clean sample after HG cut, no obvious K/p leakage - CoinTime cut helps a lot too - DG: only apply the RF cut if you have to - if get clean spectra w/o it, only HGC, then RF cut only gives an inefficiency - AU will need to look at # clean neutrons vs cut Vijay Updates ------------- Working on pion-LT package - almost ready - debugged Fortran scripts, made changes for Summer 2019 data - working on cross section function for SIMC, getting some errors as GH gave him an old version of the routine and some variable names have changed - DG: yes, some variable names in the structures have changed, see include files Next Meeting ------------ - Thur Mar 31 @ 16:00 Eastern/14:00 Regina/13:00 Pacific - KaonLT will go first