AMS - ECAL Jan 7 th -8 th, 2003 Page 1 F. Pilo - Siena University/ INFN Pisa Update on stand alone ECAL trigger S. Di Falco, M. Incagli, F. Pilo, G. Venanzoni.

AMS - ECALJan 7th-8th, 2003 Page 1

F. Pilo - Siena University/ INFN PisaF. Pilo - Siena University/ INFN Pisa

Update on stand alone ECAL trigger

S. Di Falco, M. Incagli, F. Pilo, G. Venanzoni



Trigger algorithms

Results on photon efficiency and trigger rate

Trigger robustness (against gain fluctuation, ...)

First results on electrons

OUTLINEOUTLINE



In the previous presentation a trigger algorithm has been described. This will be called “ANDAND” algorithm.

Due to the possible problems related to the low threshold values and to the rigid “AND” logic,we decided to develope a different more flexible algorithm which will be called “MIXMIX”.

Trigger algorithmsTrigger algorithms



The AND trigger algorithmThe AND trigger algorithm

thresholds on 6 superlayers:

Low energy trigger:

ANDAND of four superlayers (2,3,4,5) ++ X/Y directions cut + + mean density cut.

High energy trigger:

ANDAND of four superlayers (4,5,6,7) ++ X/Y directions cut.

Global trigger:

low energy trigger OROR high energy trigger

Superlayer Thr(MeV) Superlayer Thr(MeV)

2 20 5 50

3 50 6 400

4 100 7 230



p 1.0% (47 Hz)

Eg(GeV) 0o-10o 10o-20o

1 45% 43%

2 83% 82%

5 92% 90%

10 97% 96%

20 98% 98%

50 95% 96%

100 99% 96%

300 96% 94%

Results of the previous presentationResults of the previous presentation



New higher thresholds

Thresholds on 6 Superlayers

Thresholds logic: 3 AND 4 AND (2 OR 6) AND (5 OR 7)

New X/Y directions cut:

No density cut, No high energy trigger

2.25 10

1.6< 10

1.15< 4

bMAXNhit6+Nhit7

A new trigger algorithm: MIX logicA new trigger algorithm: MIX logic

120

100

100

Thr(MeV)

7

6

5

Superlayer

1004

1003

1002

Thr(MeV)Superlayer



New proton sampleNew proton sample

In the previous presentation the protons sample was generated

“ECAL focused”. This was not completely correct mainly because of the

secondary particles produced in detector interactions which were not

taken into account (a 30% effect was estimated). A new sample of

protons generated isotropically on the surface of the “standard box”

(390 x 390 x 390 cm3) surronding the detector was used.

The rate is defined as:

where NBOX is the number of protons generated on the box, ABOX is the

acceptance of the standard box (286.7 sr m2 ), is the flux and NTRG is

the number of events which are selected by our stand alone trigger.

BOXBOX

TRG

N

NRate A



Proton FluxProton Flux

The flux is taken from Choutko’s note ‘Over Cutoff Proton Flux…’ (May 2000) and it corresponds to the polar orbit.

1830



EpKIN

(GeV)

Flux

(Hz sr-1 m-2)

BOX

(10-4) rate

(Hz)

0.265-0.5 352 0.46 4.64 0.71

0.5-1.5 1060 0.24 7.29 1.52

1.5-2.5 502 0.47 7.76 1.01

2.5-3.5 267 0.87 6.67 0.69

3.5-4.5 160 1.37 6.26 0.55

4.5-5.5 103 1.33 3.92 0.35

5.5-6.5 70 1.57 3.14 0.26

6.5-7.5 50 2.05 2.91 0.20

7.5-8.5 36.7 1.66 1.74 0.14

8.5-9.5 27.8 1.98 1.58 0.11

9.5-10.5 21.6 2.30 1.42 0.09

10.5-11.5 17.4 2.46 1.23 0.08

EpKIN

(GeV)

Flux

(Hz sr-1 m-2)

BOX

(10-4) rate

(Hz)

11.5-12.5 14.0 2.22 0.89 0.06

12.5-13.5 11.6 2.36 0.78 0.05

13.5-14.5 9.61 2.64 0.73 0.04

14.5-15.5 7.97 2.68 0.61 0.04

15.5-16.5 6.83 2.54 0.50 0.03

16.5-17.5 5.89 3.16 0.53 0.03

17.5-18.5 4.96 2.90 0.41 0.02

18.5-19.5 4.28 3.24 0.40 0.02

19.5-20.5 3.87 3.62 0.40 0.02

>20.5 41.1 4.20 4.96 0.24

TOTAL 57.8 HzThresholds only: 482 Hz

Updated Results for the AND trigger algorithmUpdated Results for the AND trigger algorithm



EpKIN

(GeV)

Flux

(Hz sr-1 m-2)

BOX

(10-4) rate

(Hz)

0.265-0.5 352 <0.01 <0.3

0.5-1.5 1060 0.05 1.52 0.61

1.5-2.5 502 0.17 2.45 0.58

2.5-3.5 267 0.36 2.76 0.46

3.5-4.5 160 0.77 3.52 0.41

4.5-5.5 103 0.91 2.68 0.30

5.5-6.5 70 1.35 2.70 0.24

6.5-7.5 50 1.45 2.06 0.17

7.5-8.5 36.7 1.58 1.66 0.14

8.5-9.5 27.8 1.81 1.44 0.10

9.5-10.5 21.6 1.90 1.17 0.12

10.5-11.5 17.4 2.14 1.07 0.10

EpKIN

(GeV)

Flux

(Hz sr-1 m-2)

BOX

(10-4) rate

(Hz)

11.5-12.5 14.0 1.86 0.75 0.08

12.5-13.5 11.6 2.48 0.82 0.07

13.5-14.5 9.61 2.04 0.56 0.06

14.5-15.5 7.97 2.44 0.56 0.05

15.5-16.5 6.83 2.26 0.44 0.04

16.5-17.5 5.89 2.62 0.44 0.04

17.5-18.5 4.96 2.60 0.37 0.03

18.5-19.5 4.28 2.84 0.35 0.03

19.5-20.5 3.87 3.66 0.41 0.03

>20.5 41.1 3.42 4.04 0.31

TOTAL 31.8 Hz

Results for the MIX trigger algorithm: rejectionResults for the MIX trigger algorithm: rejection

Thresholds only: 188 Hz



AND vs MIX trigger algorithm: rejectionAND vs MIX trigger algorithm: rejection



In the previous presentation trigger efficiency was calculated by

requiring photon trajectory intersection with the TOF’s and the 5th ECAL

superlayer.

In the present study the photons intersecting the external ECAL columns

have not been considered (common definition with Sapinski).

New photon selectionNew photon selection



E(GeV) 0o-10o 10o-20o

1 16% 14%

2 83% 81%

3 94% 94%

4 94% 94%

5 95% 94%

10 95% 95%

20 97% 96%

50 97% 96%

100 98% 97%

300 97% 95%

E(GeV) 0o-10o 10o-20o

1 45% 43%

2 83% 83%

3 90% 89%

4 90% 89 %

5 93% 90%

10 97% 97%

20 98% 98%

50 95% 97%

100 99% 95%

300 96% 94%

OLD NEW

AND vs MIX trigger algorithm: efficiency AND vs MIX trigger algorithm: efficiency



AND vs MIX trigger algorithm: efficiency (II) AND vs MIX trigger algorithm: efficiency (II)



MIX trigger algorithm robustness: gain fluctuationMIX trigger algorithm robustness: gain fluctuation

Proton rate changes from 31.8 (no smearing) to 30.8 Hz (sigma = 20%)

Gain fluctuation was simulated by MC, applying a gaussian spread to all the PMT outputs, with sigma chosen between 2% and 20%.

fluctuation changes event by event

same fluctuation for all the events



MIX trigger algorithm robustness: gain shiftMIX trigger algorithm robustness: gain shift

Proton rate changes from 31.8 (no gain shift) to 18.9 Hz (gain shift= -30%), to 50.1 Hz (gain shift +30%)

The effect of a systematic gain shift between +30 and –30% was studied

negative gain shift positive gain shift



Proton rate changes from 31.8 (no broken PMT’s) to 29.1 Hz in the worse case.

• No broken PMT � 1 broken PMT

Effect of broken PMT’s: efficiency is the average of 20 different configurations of broken PMT’s.

• No broken PMT 5 broken PMT’s


MIX trigger algorithm robustness: broken PMT’sMIX trigger algorithm robustness: broken PMT’s




• No broken PMT � 1 broken PMT

Effect of broken PMT’s in the superlayer 3 and 4: efficiency is the average of 20 different configurations of broken PMT’s.



MIX trigger algorithm robustness: broken PMT’s (II)MIX trigger algorithm robustness: broken PMT’s (II)




• No broken HV ch. � 1 broken HV ch.

HV failure was simulated by increasing the output of some PMT’s by a factor of 10: efficiency is the average of 20 different configurations.

• No broken HV ch. 5 broken HV ch.’s

• No broken ch. 10 broken ch.’s

MIX trigger algorithm robustness: HV failuresMIX trigger algorithm robustness: HV failures



ee sample selection: 3 out of 4 TOF planesextrapolated trajectory: - do not intersect ACC - crosses 5th SL (not at edges)

ECAL fast trigger (thresh. only) to recover ECAL fast trigger (thresh. only) to recover vetoed evetoed e

TOF:TOF: 3 out of 4 TOF planes

TOF+ANTI:TOF+ANTI: 3 out of 4 TOF planes AND

NACC=0

TOF+ANTI+ECfast:TOF+ANTI+ECfast: 3 out of 4 TOF planes AND

(NACC=0 OR ECthresholds OK)

backsplash effect



Summary and OutlookSummary and Outlook

Two different algorithms were studied for the stand alone trigger:

“AND”: good low energy gamma efficiency also to 1 GeV, but low

threshold and rigid “AND” logic

“MIX”: good efficiency E≥2 GeV, higher thresholds (~100 MeV),

robusteness.

For MIX algorithm the proton rate is 30 Hz for the AMS01 proton flux

in the worst latitude condition.

Study against gain fluctuations, shifts, broken PMT’s,..., shows

sensible differences in efficiency only for energy below 5 GeV (30% in

the worse case).

Backsplash problem is well fixed.



Trigger analog circuit simulation: Jitter StudyTrigger analog circuit simulation: Jitter Study

Simulated circuit: follower + (10x) amplifier + comparator

Two tested amplifier: LMH6643, OPA690, no SQ but in the AMS list

The OPA690 amplifier requires more power (+2 mA). The overall

trigger power consumption is increased by ~1 Watt.

Using OPA690 the response delay for a small input signal is reduced

(from 60 to 25 ns)



Jitter Study: Jitter Study: LMH6643LMH6643

Input signal: 2mVDelay: ~60 ns




Jitter Study: OPA690Jitter Study: OPA690





Minimum total energy threshold of 220 MeV: good efficiency at

low energy but trigger rate sensitive to low energy proton flux

Minimum PMT threshold of 20 MeV ≳ MIP signal (~16 MeV):

could it increase fast trigger Jitter?

AND logic rigidity could reduce trigger robustness (efficiency

dependence on PMT gains and PMT failures)

Open Problems for the AND trigger algorithmOpen Problems for the AND trigger algorithm



Accettanza dell’ECAL

Calcolo geometrico. Ogni faccia conta p (flusso solo entrante):

2p (0.648 x 0.648) + 4p (0.648 x 0.1665)= 4 sr m2

Verifica ‘empirica’. Flusso isotropo su una sfera di raggio 1 m:

‘Acettanza dinamica’. Generando isotropicamente 105 fotoni da

2 GeV sulla scatola standard (3.9x3.9x3.9 m3) e guardando gli

eventi con ETOT>120 MeV sull’ECAL si trova:

AECAL(dall’alto) = 2.18 0.08 sr m2

(Corinne gives 2.2 sr m2 )

AECAL(dal basso) = 2.06 0.08 sr m2

2msr 4sfera

ECALsferaECAL N

NAA



Calcolo del rate di protoni

Per tenerne correttamente conto dei secondari si considerera’

l’efficienza non rispetto agli eventi nell’ntupla ma rispetto a quelli chiesti

al Montecarlo (TRIG card) usando come rate quello ottenuto

moltiplicando il flusso per l’accettanza della scatola standard (ABOX=

286.7 sr m2 ):

Il fattore 2 dovuto al fatto che si considerano solo particelle dall’alto e’

incluso in NTRIG (il Montecarlo genera comunque anche le particelle dal

basso ma non le scrive nell’ntupla)

BOXTRIG

trigger FlussoN

NRate A



Effetto dei secondari

Generando isotropicamente 106 protoni con spettro energetico

cosmico con 0.5<p<200 GeV/c sulla scatola standard (3.9x3.9x3.9 m3)

si trova:

ApECAL(dall’alto) = 2.72 0.08 sr m2

ApECAL(dal basso) = 2.07 0.08 sr m2

L’eccesso del 35% nel flusso di particelle dall’alto e’ da imputare ai

secondari

Per tenerne conto si considerera’ l’efficienza non rispetto agli eventi

nell’ntupla ma rispetto a quelli chiesti al Montecarlo usando come rate

quello ottenuto moltiplicando il flusso per l’accettanza della scatola

standard

AMS - ECAL Jan 7 th -8 th, 2003 Page 1 F. Pilo - Siena University/ INFN Pisa Update on stand alone ECAL trigger S. Di Falco, M. Incagli, F. Pilo, G. Venanzoni.

Documents

ecal trigger

ams ecal

trigger rate

trigger algorithm slide

generated ecal

ecal dallalto

trigger algorithms

new trigger algorithm