TRIGGER PRIMITIVE RATES IN SIMULATION Karl Warburton Iowa State University 20 th June 2019 1
MOTIVATION FOR LOOKING INTO THIS • At a meeting on 15 th October, Phil showed some Trigger Primitive rates from ProtoDUNE data. Slides here, and some screenshots from this presentation are shown below. • • It was noted that it would be interesting to see how this compares with simulation. • Unfortunately, the most similar samples are in the 1x2x6 geometry, not ProtoDUNE, so the plots that I’m about to show have to be taken with a slight pinch of salt. 2
WHAT I’M SHOWING TODAY / TOOLS USED • Apologies for a couple of very terse slides… • In case anyone wants to look at any of the work • Using the “ Tools for Trigger Primitive Studies ” framework, which is described in detail here. presented, all of my code resides in the various subdirectories contained in • Assuming that this is the most up-to-date version of code. I am using dunetpc released on 12 th April ( v08_15_00 ). /dune/app/users/karlwarb/LArDevelop/WorkArea_DAQ/TrigPrim_ProtoDUNE/Simulation-1x2x6/ • Used the default noise levels, and the default trigger primitive algorithm finders. • I did add a number of different trigger primitive threshold modules though (seen right). • I don’t know what ADC values the different thresholds correspond to though. • Did anyone come up with a conversion matrix for them? 3
WHAT I’M SHOWING TODAY / TOOLS USED • The TrigPrimAna module identifies the generator • Used outputs of samples the DAQ group requested in responsible for each Trigger Primitive. MCC11, here. • Be that noise, a radiological mode, or MARLEY, etc. • Specifically the MARLEY sample (containing radio bkg) and • Can therefore “just” run the module, and then after the 5 times Radio Bkg only sample. getting the total number of TrigPrim’s per event, • I’m only showing plots from the baseline radiological rate calculate an expected rate for continuous running. today, though I can quickly spin up the other sample. • The FlatTrees that are written out by theTrigPrim Ana • Not showing any MARLEY numbers here, but NB: module are very large, so this represents only 10% of • Sample has 3 interactions per event, so need to scale. the full stats, but the flat tree is still 80 GB! • Obviously not going to be having SN interactions in every • Could very well reduce the amount of info stored so that I drift window, so it’s more an indication of what we can can process the full sample. expect if a SN does occur. 4
NUMBER OF ARGON 39 TRIGGER PRIMITIVES FOR DIFF THRESHOLDS Hits_Per_TPC_0_Ar39 Hits_Per_TPC_0_Ar39 • Forgot a legend, but; 3 ´ 10 Number of events 100 • The lowest threshold (800) is black, mean of around 70. 80 • The highest threshold (3200) is reddy-brown, mean < 10. 60 40 Hits_Per_TPC_1_Ar39 Hits_Per_TPC_1_Ar39 • Argon39 is by far the highest number of Trigger Prims, and ´ 3 10 Number of events 20 we can’t really tell much from this plot, so won’t be 100 0 20 40 60 80 100 120 showing any more of these plots. Number of Hits 80 • All you can tell is that the number of trig prims is reduced 60 quite sharply by an increasing threshold. 40 20 • T op Plot isTPC 0, Bottom Plot isTPC 1. 0 20 40 60 80 100 120 140 Number of Hits 5
NUMBER OF ARGON 39 TRIGGER PRIMITIVES FOR DIFF THRESHOLDS Hits_On_All_TPCs_Ar39 Number of events • Forgot a legend, but; 4 10 • The lowest threshold (800) is black, mean of around 1700. 3 10 • The highest threshold (3200) is reddy-brown, mean ~30. 2 10 Hits_On_All_TPCs_Ar39 Hits_On_All_TPCs_Ar39 10 80000 Number of events • Again, can’t really tell much other than the obvious, 70000 1 however the separation is quite cool. 3 2 10 10 60000 Number of Hits 50000 • Note how large the variation in NTrigPrims is for the lower 40000 thresholds (the RHS of the plot). 30000 • Hints as to later, doubling the threshold from 800 (black) to 20000 10000 1600 (green) reduces the rate by a factor of ~4. 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Number of Hits 6
CALCULATING THE TRIGGER PRIMITIVE RATE No error bars on this plot, as taking the RMS of • TriggerPrimRate_AllTPC TriggerPrimRate_AllTPC the distribution seems to give too large errors Ar39 Rate (Hz) 6 10 (at least aesthetically). Unknown 5 10 APA Note that this is the whole detector, so divide by • 4 10 24 to get the per TPC rates. CPA 3 10 Neut • Ar39 dominates at all thresholds, noise is 2 10 removed quite quickly, and there were never any Kryp 10 Trigger Primitives for the Polonium generator. Radon 1 Ar42 • 1 event = 4492 ticks (500 ns) = 2.246e-3 s. - 1 10 800 1000 1200 1400 1600 1800 2000 2400 2800 3200 T otal of 1e5 evts à 224.6 s in this study. • Threshold 7
TriggerPrimRate_TPC0 TriggerPrimRate_TPC0 TriggerPrimRate_TPC2 TriggerPrimRate_TPC2 Ar39 Ar39 Rate (Hz) Rate (Hz) TRIGGER 4 4 10 10 Unknown Unknown PRIMITIVE APA APA 3 3 RATES PER 10 10 CPA CPA TPC 2 2 10 10 Neut Neut 10 10 Kryp Kryp • Busy slide, but interest. Radon Radon 1 1 to have all together. Ar42 Ar42 - - 1 1 10 10 800 1000 1200 1400 1600 1800 2000 2400 2800 3200 800 1000 1200 1400 1600 1800 2000 2400 2800 3200 Threshold Threshold • Most gen’s show TriggerPrimRate_TPC1 TriggerPrimRate_TPC1 TriggerPrimRate_TPC3 TriggerPrimRate_TPC3 same pattern per TPC Ar39 Ar39 Rate (Hz) Rate (Hz) 4 4 10 Unknown 10 Unknown as for full TPC, and APA APA the rates are very 3 3 10 10 CPA CPA stable. 2 2 10 10 Neut Neut • However, the CPA 10 10 Kryp Kryp gen only makesTrig Radon Radon 1 1 Ar42 Ar42 Prims in oddTPCs… - - 1 1 10 10 800 1000 1200 1400 1600 1800 2000 2400 2800 3200 800 1000 1200 1400 1600 1800 2000 2400 2800 3200 Threshold Threshold 8
SUMMARY / NEXT STEPS • Open questions: • Not shown: • Need to have a think about the CPA only making TrigPrims • The average number of Marley Trig Prims per event (3 in odd TPCs. Maybe the radiological group expects this? interactions) goes from 35 (Thr: 800) to 21 (Thr: 3200). • Fix the error bars on the rate plots, will be interesting to Raw numbers of Trig Prims per TPC for anything other than • see what the event to event variation is around the mean. Argon39, as above, I don’t think we can tell much though. • How do the rates scale when radiological background is 5 • Large scale steps: times larger? • Josh expressed a need to reproduce Phil’s plots for “higher- • Niggles: quality runs” as these were made for runs just after the • If we want to see more plots like Slides 5/6, I’ll have to stack detector turned off. Did that ever get done? the hists or do a slightly better binning. • Don’t have any cosmics in here. Did we request any cosmic • Do we expect Polonium to not produce any TrigPrims? samples? What generator do we use? • Conversion of Trigger Threshold into ADCs. 9
BACKUP SLIDES 10
Recommend
More recommend