This page contains discussion of the trigger setup of the one of 2014's winning experiments: Pion Decay experiment from the team Odysseus' Comrades.

Acceptance calculations

Acceptance Numbers: Percentage of particles (electrons or muons) that were observed in detector to the whole population (pions?) that decayed after SC0, before Calorimeters

Beam Energy Electron Muon
1.0GeV 26.52 84.8
2.0GeV 47.03 100.0
3.0GeV 61.41 100.0
4.0GeV 71.37 100.0
5.0GeV 78.26 100.0
6.0GeV 83.16 100.0
7.0GeV 86.63 100.0
8.0GeV 89.21 100.0
9.0GeV 91.11 100.0
10.0GeV 92.57 100.0

Acceptance Numbers: Percentage of particles (electrons or muons) that were observed in detector to the whole population (pions?) that decayed after SC0. This is a more important number, since the ones that decayed after SC0 are the pions that trigger the DAQ. It assumes calorimeter at 10m from SC0.

Beam Energy Electron(%) Muon(%)
1.0GeV 3.4 10.90
2.0GeV 3.19 6.77
3.0GeV 2.83 4.59
4.0GeV 2.49 3.48
5.0GeV 2.19 2.80
6.0GeV 1.94 2.34
7.0GeV 1.74 2.01
8.0GeV 1.57 1.76
9.0GeV 1.43 1.56
10.0GeV 1.31 1.41

Triggers and Measurements

Trigger Units:

  • CH0 -> Set above pion mass (accepts pions muons electrons)
  • CH1 -> Set below pion mass (accepts muons electrons)
  • SC0 (T9)
  • SC1 (Muon)
  • SC2 (Pion/Muon)

Trigger Types

Pion Trigger

BL4SPionTrigger.jpg

Possible Measurement:

  • from data, find a sample of pion to muon decays, N( muons, observed).
    • the muons are identified by a muon signal in the LG and a hit in SC1 (and SC2?)
  • correct this for acceptance. We then have the number of pions that decayed into muons after SC0: N( SC0, muons)
  • from data, find a sample of pion to electron decays, N(e, observed).
    • the electrons are identified by a signal in the LG
  • correct this for acceptance. We then have the number of pions that decayed into electrons after SC0: N( SC0, electrons)
  • BR(e+) / BR(muons)= N( SC0, electrons) / N( SC0, muons)
  • the number of pions at SC0 = # events in the datafile = N(SC0, pions) during livetime
  • BR(e+) = N(e,electrons) / N( SC0, pions)
The main tool for identication of electrons and muons is the lead glass (LG) detector, using a cut in the energy spectrum or a fitting method. The difficult part is to find a sample of pion to electron decays. The acceptance plots show that ~2% of the pion (triggers) decay between S0 and the lead glass calorimeter. Of those 1/10000 are decays to electrons. Therefore, in 10**6 events we can expect a few decays to electrons. This corresponds in best conditions to 30 minutes of data recording. The electrons have a spread in energy (fig) and may be partially absorbed by the frame of DWC2. In addition, in the energy spectrum in the lead glass detector there is a hadron tail. It's not clear that an electron sample can be found. To address this problem, another trigger is proposed as described in a section below.

Electron trigger: Selects pions decaying into electrons

The idea behind the electron trigger is based on a simple observation: the electrons from pion decays between S0 and the LG are absorbed in the LG. Therefore, a scintillation counter placed behind the LG detecting pions and muons and used as a veto, allows to "remove" pions and muons from the pion triggers and thus select the pion to electron decays - or at least to provide an event sample with a significantly increased number of electrons.

BL4SElectronTrigger.jpg

  • it is seen that the e+ trigger is an add-on to the pion trigger, it doesn't require any recabling of the pion trigger. The e+ trigger signal is interesting in itself without being used as a trigger. The scaler counts of e+ should be compared to the count of the pion trigger - it may be higher since it includes muons.

  • In order to switch from the pion trigger to the electron trigger two cables changes are required: an output from CoPi goes to a second coincidence CoTwo. Move it to the input of CoE+. Connect an output from CoE+ to CoTwo. Or the output from CoE+ could go into another input of CoTwo, then no re-cabling at all is required.

  • the veto from SC2 should be before the pion(trigger) signal which may require inserting a delay in the pion(trigger). The pion(trigger) is already delayed (a bit) by the pion trigger logic but to avoid or at least minimise the delay, the cable from SC2 should be as short as possible. In more details: the pion(trigger) is delayed due to the coincidence unit CoPion (~10 ns). Due to the time of flight, the signal from S2 is delayed ~30 ns(10m) but if the cable is shorter by 10m this delay is compensated. Thus switching from the pion to the electron trigger may not require any timing adjustments.
  • additional hardware requirements:
    • a scintillator (SC2), and a support. Actually, two scintillators may be useful, the OR between them would increase the efficiency. We have two ISOTDAQ scintillators in the lab bdg.40
    • a 'short' cable from SC2 to the control room (30m?). Actually a short cable would also be useful for SC1.
    • coincidence, discriminator, TDC channel, two scaler channels
Possible Measurement:
  • from data, find a sample of pion to electron decays, N(e, observed).
  • correct this for acceptance. We then have the number of pions that decayed into electrons N(pion, electron) after SC0: N( SC0, electrons)
  • correct for deadtime, in this case it's negligible since decays into electrons are rare.
  • we know the number of pions (triggers) at SC0: N( SC0, pions). This is the scaler count PionTriggers.
  • BR(e+) = N( SC0,electrons)/ N( SC0,pions)
Comment: as we discussed briefly, one could make a downscaled veto trigger, so that only a KNOWN fraction (e.g. 95%) are used to veto. For eaxmple using a coincidence between a regular pulse train and the veto signal one can remove a fraction of the signals.

Comment: not really a trigger comment. It's not clear that DWc2 is really useful? IF not it should be removed since it will be in the way for a lot of electrons??

Electron/muon trigger: A simpler version

The electron trigger selects pions that decayed to electrons before the LG. The SC1 scintillator selects muons after the muon filter. The logical OR between them then defines a sample of pions that decay to electrons before the LG and pions that decay to muons before SC1.

BL4SSimpleElectronOrMuonTrigger.jpg

Possible Measurement:

  • from data, find a sample of pion to muon decays with pions decaying BEFORE the LG. Events must have a hit in the SC1 (all triggers should have ..) AND a muon signal in the LG.
  • correct this for acceptance. We then have the number of pions that decayed into muons after SC0: N( SC0, muons)
  • from data, find a sample of pion to electron decays, N(e, observed) like in the case of the electron trigger.
  • correct this for acceptance. We then have the number of pions that decayed into electrons after SC0: N( SC0, electrons)
  • BR(e+) / BR(muons)= N( SC0, electrons) / N( SC0, muons)
  • the number of pions at SC0 = # events in the datafile = N(SC0, pions) during livetime
  • BR(e+) = N(e,electrons) / N( SC0, pions)

Electron/muon trigger: Selects pions decayed into electrons/muons

The idea behind the electron/muon trigger is similar to the electron trigger. A scintillation counter placed behind the muon filter detecting muons and used as a veto, allows to "remove" muons from the SC1 trigger and thus select the pions, only, after the LG. If this signal is used, in turn, as a veto for the pion trigger at SC0, we are left with electrons and muons from pion decays at SC0. This trigger may also include background muons from the pion trigger at SC0.

BL4SElectronOrMuonTriggerV2.jpg

  • it is seen that the e+/muon trigger is an add-on to the pion trigger, it doesn't require any recabling of the pion trigger. The e+/muon trigger signal is interesting in itself without being used as a trigger.

  • In order to switch from the pion trigger to the e+/muon trigger two cables changes are required: an output from CoPi goes to a second coincidence CoTwo. Move it to the input of CoE+Muon. Connect an output from CoE+Muon to CoTwo. Or , simpler, the output of CoE+Muon goes into another input of CoTwo.

  • the veto from SC1 should be before the SC2 signal which may require inserting a delay in the SC2 signal (only the discriminator output cable from SC2 to CoPiLg). This could be avoided if the cable from SC1 is even shorter than the cable from SC2. The time-of-flight from SC2 to SC1 is about 7 ns so a reduction in cable length of 2-3 m would be sufficient.

  • additional hardware requirements (SC1 is already there ..):

    • a 'short' cable from SC1 to the control room (30m?).
Possible Measurement:

  • from data, find a sample of pion to muon decays, N( muons, observed).
  • correct this for acceptance. We then have the number of pions that decayed into muons after SC0: N( SC0, muons)
  • from data, find a sample of pion to electron decays, N(e, observed).
  • correct this for acceptance. We then have the number of pions that decayed into electrons after SC0: N( SC0, electrons)
  • BR(e+) / BR(muons)= N( SC0, electrons) / N( SC0, muons)
  • the number of pions at SC0 = # events in the datafile = N(SC0, pions) during livetime
  • BR(e+) = N(e,electrons) / N( SC0, pions)

Direct Electron trigger: Selects pions decaying into electrons based on Lead Glass signal (preliminary).

It has been proposed to use the LG detector for triggering on electrons. To trigger on the total signal from the LG is complicated. A simpler trigger derived from four LG blocks would be simpler. A minimum of four is required to contain the electron signal and provide a reasonable acceptance. The latter could be about 25-30% of the one corresponding to the full surface of the LG, depending on beam momentum.

BL4SDirectElectronTrigger.jpg

The analogue signals from four blocks are split by an (active) analogue fanout (LeCroy428F?). Four outputs go (unchanged) to the QDC. Four others are input to a FIFO where the analogue sum is formed. This requires precise timing of the inputs to the first FIFO. An output goes to a discriminator with a high threshold to produce a signal for electrons (possibly including a tail of hadrons). This is put into a coincidence with the Pion Trigger to produce an event trigger. This trigger could be an alternative to the rather complicated electron/muon trigger

  • additional hardware requirements:
    • analogue fanout and fanin (LeCroy 428F?). Coincidence channel. Scaler channel.

Possible Measurement:

  • from data, find a sample of pion to electron decays, N(e, observed).
  • correct this for acceptance calculated from the surface of the four trigger LG blocks used in the trigger . We then have the number of pions that decayed into electrons N(pion, electron) after SC0: N( SC0, electrons)
  • correct for deadtime, in this case it's negligible since decays into electrons are rare.
  • we know the number of pions (triggers) at SC0: N( SC0, pions). This is the scaler count PionTriggers.
  • BR(e+) = N( SC0,electrons)/ N( SC0,pions)

Deadtime

The discussion below the figure is probably too complicated and obsolete.

From the scaler counts we know the number of triggers of each type: XTriggers where X = Pion, Electron, ElectronMuon or SimpleElectronMuon depending on which trigger is used. Then

LiveTime (%) = EventTriggers/!XTriggers and * How do we calculate the LiveTime in seconds - bursts *!LiveTime(secs) = LiveTime(%) * MilliSeconds/1000

DeadTime (%) = 100 - DeadTime(%)

BL4SDeadTimeTwo.jpg

  • the first question is: do we need to know the deadtime during a run?
  • can the burst structure be measured i.e. is there a precise burst signal available?
  • if we assume that the bursts are identical and regular in time, t(active)/t(real) = 1.4s/6s (?????????)
  • if the value of the ms counter is recorded is recorded with every event, can the exact burst structure or 'active beam time' t(active) be computed from the data in a run?
    • for example: assume that there is at least 'some' events per burst and that the minimum time between bursts is known
    • find the events in the first burst(check on count 'close' in time).
    • find the events in the next burst ....
    • compute the number of bursts in the run
    • assuming that the burst time t(burst) is constant : t(active) = sum(t(burst))
      • if there are 'many' events per burst, t(burst) = t(last event in burst) - t(first event in burst)
If we know the active beam time then we can compute the deadtime:

# counts with active beam N(burst) = N(pulses) * t(active) / t(real)

deadtime = N(busy)/N(burst)

We should check that these signals are sent to scaler channels.

The deadtime measurement allows to compute the real number of events in a run from the observed number: Nreal = Nobserved/!LiveTime e.g NPionsSC0 = NpionsObserved/LiveTime = NEventsInDatafile/!LiveTime

Scaler and Detector naming conventions

This section should probably not be here .. The purpose is to define more consistent naming conventions related to scalers and detectors and to define the corresponding scaler channels in the V560. In cable diagrams, signal plots and Twikis names are used in a somewhat inconsistent and sometimes misleading way. The logical names in the table should be 'intuitively' understandable and could be used in outputs of scaler information in monitor programs, run summaries etc.. The 'physical' name are used in cabling diagrams, signal timing diagrams, Twiki trigger pages etc.

Logical names Physical names Scaler channels Comment Config File Tags for channels of scalers
EventTriggers CoTwo (Main Trigger) 0 # CORBO triggers = # Events in the datafile N_EVENT_TRIGGER_CH ; N_EVENT_CH
DetectorTriggers CoDetector 1 Events coming from the selected detector trigger N_DETECTOR_COINCIDENCE_CH ; N_DETECTOR_CH
T9Scintillator SC0 2 # hits in SC0 N_T9_SCINTILLATOR_CH ; N_SCINTILLATOR0_CH
CerenkovPionMuonElectron C0 3 Higher pressure to accept pions, muons, electrons N_CHERENKOV0_CH
CerenkovMuonElectron C1 4 Lower pressure to accept muons, electrons N_CHERENKOV1_CH
MuonAfterFilter SC1 5 Hits in the muon scintillator SC1 N_MU_AFTER_FILTER_CH ; N_MUON_SCINTILLATOR_CH ; N_SCINTILLATOR1_CH
PionMuonAfterLeadGlass SC2 6 Hits in the scintillator after the leadglass (SC2) N_PI_MU_AFTER_LG_CH ; N_SCINTILLATOR2_CH
PionTriggers CoPion 7 # pions at SC0 N_PI_TRIGGER_CH
ElectronTriggers CoE+ 8 Initial particle: Pion Final Particle: Electron N_E_TRIGGER_CH
SimpleElectronMuonTriggers OrE+Muon 9 Final Particle: Electron or Muon N_SIMPLE_E_MU_TRIGGER_CH
ElectronMuonTriggers CoE+Mu 10 Initial particle: Pion Final Particle: Electron or Muon N_E_MU_TRIGGER_CH
MilliSeconds   11 real time counter in ms N_MILLISECONDS_CH ; N_SECONDS_CH

-- CenkYildiz - 20 Aug 2014 -- SaimeSarikaya - 28 Aug 2014

Topic attachments
I Attachment History Action Size Date Who Comment
JPEGjpg BL4SDeadTimeTwo.jpg r1 manage 28.6 K 2014-08-25 - 15:31 JorgenPetersen  
JPEGjpg BL4SDirectElectronTrigger.jpg r3 r2 r1 manage 38.3 K 2014-08-31 - 15:40 JorgenPetersen  
JPEGjpg BL4SElectronOrMuonTriggerV2.jpg r1 manage 44.4 K 2014-08-25 - 17:37 JorgenPetersen  
JPEGjpg BL4SElectronTrigger.jpg r3 r2 r1 manage 33.2 K 2014-08-22 - 18:06 JorgenPetersen  
JPEGjpg BL4SPionTrigger.jpg r3 r2 r1 manage 36.6 K 2014-08-28 - 15:39 JorgenPetersen  
JPEGjpg BL4SSimpleElectronOrMuonTrigger.jpg r1 manage 20.3 K 2014-08-27 - 12:51 JorgenPetersen  
Edit | Attach | Watch | Print version | History: r32 < r31 < r30 < r29 < r28 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r32 - 2015-07-04 - JorgenPetersen
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    BL4S All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback