FVTX Trigger

Toru Nagashima, Yomei Fukushima (Rikkyo University)Shoichi Hasegwa

(J-Parc)Itaru Nakagawa

RIKENand

FVTX Group

1

MOTIVATION

2

CMS Reported “Ridge” feature in pp,pA

Slide from Wei Lei, Rice University3

High Multiplicity Events is the key!

4

IHigh multiplicity trigger in FVTX is to be developed for Run15!

Slide from Wei Lei, Rice University

Experiment pA, dA pp

CMS ✔ ✔

ATLAS ✔

PHENIX ✔

5

PPG161

PPG Chair: Shengli Huang (Vanderbilt) PPG Members: PPG Members: John Chin-Hao Chen(RBRC), ShinIchi Esumi(Tsukuba), Jiangyong Jia(Stony), Anne Sickles(BNL), Julia Velkovska(Vanderbilt), Paul Stankus (ORNL)

In preparation for PRL

6

Observed v2 in dAu

Minimum bias : pp

7

Elementary process (not flow)

He3-Au V2MPC South MPC North

RIDGE is observed for centrality < 20% in MPC 8

Key Features to observe Ridge

• 0 ~ 5% centrality => 20 ~ 40 tracks / rapidity unit• 80 Mevents Cental dA Events in Run8

9

BBC High Multiplicity Trigger for HeAu

10

Why not BBC High Multiplicity Trigger works on pp?

11

Num

ber o

f Tra

cks

in P

C

BBC Charge

Ridge

Doesn’t show strong correlation between BBC and central arm activities

HeAu 200 GeV pp

Clear correlation btwn central arm activities and BBC multiplicity

-> Works as trigger -> Doesn’t Works as trigger

Trigger sensitive to high multiplicity

• Central arm : trigger on the number of 2x2 fired cells in the EMCal -> interfere with other electron trigger.

• FVTX Trigger (in preparation)

12

FVTX trigger is the best if (mini-QGP) is seen in FVTX rapidity region. So far the best trigger for central arm high multiplicity in pp.

(Need correlation study between FVTX and Central arm.)

Required Rejection Power# of Tracks / arm > 0 > 10 > 20 > 30 > 40Fraction in MB 100% 10% 0.7% 0.04% 0.008%Rates in MB 1MHz 100kHz 7kHz 400Hz 80HzRequired Rejection Power 1 10 142 2,500 12,500

13

# of Tracks in FVTX / arm @ 510GeV pp

Run13 Data Possible band width for FVTX Trigger

Key Features to observe Ridge

• 0 ~ 5% centrality => 20 ~ 40 tracks / rapidity unit• 80 Mevents Cental dA Events in Run8

14

• FVTX North and South separate trigger bits• 100Hz per arm (Total 200Hz)• 200Hz x 3600s x 24h x 7days x 9 weeks x 30%

DAQ time ~ 300M high multiplicity events

Goal of Run 15

-> Purity of the trigger is the issue

FVTX TERMINOLOGY AND FVTX TRIGGER

15

Debugging FPGA @ BNL Test Bench

September, 2014

16

FVTX

Wedge

17

18

• At least one hit in 3 out of 4 stations is required within a half sector ((Df=7.5o). -> TrackFlag.

• If there are more than two tracks within a given sector (Df=15o), the second one won’t be counted.

We Request Two Trigger bits (one per arm) for FVTX trigger.

AND Limited to 1bit

FEM -> FEM-IB Cable

19About 30 cm long.

FEM Board

20

Test Pin OUT

Serial Wire

To FEM-IB Test-Pin

21

FEM -> FEM-IB Installation

22November 5th, 2014

FVTX Terminology

23

Station-0Station-1

Station-2Station-3

Wedge

Sector

Wedge

FEM

Cage

FVTX Trigger Parameters

24

Station-0Station-1

Station-2Station-3

Wedge

Sector

Wedge

FEM

Cage

# of hits/wedge

# of active wedges out of 4 stations

Active Sector (AND/OR)

# of active FEMs

Arm

Active Cage (AND/OR)

Trigger Concept

Multi Track Multi Hit

How well we can trigger multi track events by multi hit algorithm?

25

Trigger used in CMS

Multi-Track and Multi-Hit Triggers

26

Algorithm Configuration Advantage Disadvantage

Multi-Track

3 active hits out of 4 stations

More sensitive to Track limited to 12 tracks/cage

Multi-HitTotal Sum of

hits in 4 stations > n

Can apply high threshold per wedge hits Less proportional to tracks

Parameter Conditions for this study

27

Algorithm Configuration Advantage Disadvantage

Multi-Track

3 active hits out of 4 stations

More sensitive to Track limited to 12 tracks/cage

Multi-HitTotal Sum of

hits in 4 stations > n

Can apply high threshold per wedge hits Less proportional to tracks

OR

Run InformationRun 13 510 GeV pp

Run# 387661

DAQ Time 48 minutes

Raw BBC Rate 1.2MHz

Scaled BBC Rate 51 Hz

Total MB Samples 147k events

MULTI-TRACK TRIGGER ALGORITHM#of Active FEMs = 6,9,12 / cage

Multi-Track Mode (South)

Number of Tracks/arm Number of Hits/arm

Number of active FEMs >=6 / cage

MULTI-HIT ALGORITHM

MultiHit (South)# of Active FEMs @ Cage > =6

Multi-Track vs. Multi-Hit Algorithm

MultiTrack

MultiHitLow track events are more suppressed in Multi-Track than Multi-Hit Mode

# of Active FEMs @ Cage >= 9

FVTX Trigger Performance

1 10 100 1000 100000

0.2

0.4

0.6

0.8

1

1.2

Multi-Track AlgorithmMultiHit (Track/Cage>6)Multi-Hit (Track/Cage>9)

Rejection Power

Trig

ger E

ffici

ency

(#Tr

ack/

arm

>25)

Conclusion

• Turn on is good against the hit multiplicity, but cannot avoid low #track leak for the track multiplicity.

• So far, satisfactory rejection power and efficiency is not achieved in data analysis.

• Further study to be made to seek for better performance.

• If RP isn’t sufficient, the trigger needs to be prescaled. However, the FVTX trigger still provides orders of magnitudes more high multiplicity samples than MB.

40

DIFFERENCE BETWEEN ONLINE AND OFFLINE TRACK

41

Difference in Track Criteria

This track will be recognized as track in both online and offline.

A

FEM track frag

Offline track

FEMi

B

C

B

This category won’t be picked up by the trigger -> cause of survival tracks

Difference in Track Criteria

This track leaves only 2 hits in FVTX and won’t satisfy ¾ station criteria in the trigger

A

FEM track frag

Offline track

FEMi

B

C

C

Difference in Track Criteria

• This differences can cause small fraction survives FEM track frag>x trigger condition

A

FEM track frag

Offline track

FEMi

FEMi+1

B

CA

CC -> B

This category won’t be picked up by the trigger -> cause of survival tracks

Possibly remove the border within FEM FPGA

FVTX Alignmenty-axis

x-axis

FVTX Event Display

track

St-0 St-1 St-2 St-3

Wedges are slightly staggered in phi direction to gain phi resolution on purpose

This mismatch of phi angle between stations enhances “category C”

Low # Track Leak Cause Summary1. Online trigger won’t recognize track which

crosses boarder between sectors2. Online trigger won’t have VTX hit information.3. FVTX stations are staggered by a little bit in phi

direction

47

No trivial solutions without significant revise in hardware configuration

BACKUP

48

He3-Au: BBC Charge vs. #Track in PCN

umbe

r of T

rack

s in

PC

BBC Charge

Ridge

49 -> About 20 tracks in PC in the centrality < 20% in HeAu

Multiplicity vs. rapidity

50

PHOBOS Phys. Rev. C72, 031901

d-goingAu-goingv2 smallerLarger

Multiplicity vs. rapidity

51

He3-goingAu-goingv2 smallerLarger

He3Au

~ 60

~ 25

On average ~40 tracks is a guideline to observe v2

Run15 pp Goal

• FVTX North and South separate trigger bits• 100Hz per arm (Total 200Hz)• 200Hz x 3600s x 24h x 7days x 9 weeks x 30%

DAQ time ~ 300M high multiplicity events• (80 Mevents Cental dA Events in Run8)

52