1Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

Fast Timing Counters for FP420/CMS and Test Beam Particle ID

Mike Albrow, September 14th 2006

FP420 (Forward Protons at 420m from CMS/ATLAS) R&D project

p pH

Measure p’s very well M(H), J=0?, CP = ++, Hgg coupling...Difficulty: SM-H(120-160) ~ 1-10 fb.Need high luminosity (~ 30/fb) with multiple interactions.Pile-up: p’s from different events.

Precision timing on p’s can reduce this:iff from same interaction.

( ) 10ps ( ) 2.1mmt z ( )[interactions] 55 80mmz

Factor 20 – 30 reduction in pile-up B/G, + kinematic constraints.

2Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

For FP420 we have tested two prototype designs.Both use Cerenkov light – prompt – GASTOF and QUARTIC.

pMCP-PMT

45-MIRROR

THIN WINDOW

1 atm. C4F8O

1 m.

GAS-Time-Of-Flight

Aperture ~ 10 x 25 mm

This is what I am suggesting to use for TOF ID in MTEST.

QUARTIC (QUARtz TIming Cerenkov)Quartz bars 6mm x (6 – 1.5)mm x 20mmwith air light guides to MCPMT pixels.8 measurements on a track. Can measure2 separate tracks >3mm in y.

Used area 10mm x 25mm~ 60mm quartz so thick for MTEST.

All light arrives within 3 ps

3Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

GASTOF for prototype tests t958

Univ. Louvain

For MTest: 1m. longThin mirror/windows

4Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

In Mtest (until this evening!):

GASTOF 1 GASTOF 2 QUARTIC 1QUARTIC 2

5Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

It’s (10ps) been done!The challenges arethe electronics and geometry for FP420.SLAC group (Va’vra)got 30ps for single p.e.

6Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

Photodetector

We concentrate on Burle 2” x 2” MCP PMT (Micro-channel plate PMT)Hamamatsu have higher performance with 10 um channels, but area smaller.Burle has 25 um channels (makes it slower) but are developing 10 um.

Burle have given us tubes to test. 25um pores and one 10 um pore tube.Univ. Louvain bought 2 tubes for GASTOFs M Test.These have an 8 x 8 array of 6mm pixels ( 48mm x 48mm (2”))For GASTOFs one output ... couple pixels “isochronously”

7Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

Cerenkov light in C4F8O (Octafluorotetrahydrofuran, 2.5)

2ch 2 2

2pe

2pe

pe

1 1cosθ = ; sin θ 1.0

βn β n

n(λ) (C4F8O) = 1.001375 for λ 400nm - 650nm

N 490 ( . .) (cm) sin θ

For L = 1m., Q.E. = 20% N 9800 sin θ

N ( 1) 27

Q E L

... at atmospheric pressure.

At 2 atm. K/p threshold is 6.5/12.5 GeV/c.Asymptotic #pe = 54

non-flammable and non-toxic

8Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

e

1

2.( /100 )t ns

9Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

We must have a good reference signal free of jitter between stations. This is as important as the detectors themselves.

Temperature control? Return path control?

Need to discuss with RF/clock experts. Sophie Baron is contact at CERN for LHC timing issues.

Assuming jitter problem solved, calibrate with (p-independent).

I have not compared abilities with conventional (P.H. or RICH or DIRC etc) Cerenkovs.This is a new approach which (a) meshes well with CMSapplication (b) Could be developed (Henry!) to 1 ps

M(π), M(K), M(p)

2 2cm detectors before m ...π/μ at 10 GeV/c

and π/K at 50 GeV/c!

10Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

CMSor

ATLAS

LHC-RF (400 MHz) sharp regular pulse

“TDC” “TDC”

QUARTIC/GASTOF

QUARTIC/GASTOF

START START

STOP STOP

Optical Fiber Optical Fiber

Comparator(control)

May not be necessary, but can testfor L-R propagation time

differential changes

E W

BPM(optimized for timing)

BPM(optimized for timing)

Tell where p is w.r.t. bunch centroid. (few mm/70mm)Fine correction on p_incident. Compare with sum time from E Wt +t

p p

75.5 mmz 75.5 mmz

A possible reference time scheme (LHC):

(250 ps)

Single mode?

CENTIME

?

FO

11Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

Three Gastof’s : redundancy, cross-checks, improved resolution(3-point fit to “timetrack”)

Electronics, Set-up for MTest

MCPPMT

HVv.stable

no ripple

ATTENUATOR?

AMPLIFIER?

CONSTANTFRACTION (?)

DISCRIMINATOR

ADC

TDC<=25ps/bin

OPTICALREFERENCE

TIMEOPTICAL

toELECTRICAL

LOCAL CENTRAL

Only optical fibers

TDC bin/4differences

Temperature controlledStabilized Power SuppliesFaraday cage?Short straight SMX cables, minimize connectors

, , , ,e K p

light

12Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

,L it ,R itL R

,L it

,R it

tz

This is ambitious, not for LHC Day 1, but P/U worse in 2010!This is ambitious, not for LHC Day 1, but P/U worse in 2010!

This can identify (in principle) interactions that have particles

in both L and R detectors. All are background for p+H+p!

Pads 1cm ×1cm

with 10 20 ps

covering (e.g) 3.9 | | 4.9

at (5.0?)7.5m

t

z

z from centraltracks

13Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

Fast Timing Layer Detectors

~ Existence proof:Burle 48mm x 48mm MCP PMT (Micro-channel plate PMT) as for QUARTICDeveloping 10 micron channels, approaching 10 ps.

Tile the plane with these

QUARTZPLATE

MCP-PMT

PARTICLES Approximate numbers:~ 250 each side, perhaps 1000 1cm x 1cm padsExpensive solution (~1M$?) but cost could drop.

Questions: Simulation, acceptance, background crapand especially RADIATION HARDNESS. Another solution?

14Mike Albrow Sept 14th 2006Time-of-Flight with GASTOFs im Test Beam

SUMMARY

High precision timing (<~ 10-20 ps) has important future roles:

It is essential for pile-up rejection in pp p+H+p. p+WW+p at LHC

A very similar set-up can do particle I.D in beams (MTest Upgrade)

<2 ps, large areas open up new possibilities (I.D., Gaps in spacetime)

MCP-PMT is probably best fast photo-detector for now. Later??