View
9
Download
0
Category
Preview:
Citation preview
FAST
FAST:
an experiment for a high
precision muon lifetime (and GF )
measurement
LucaMalgeriUniversity of Geneva
July 3, 2002 - νFact02, London Run 7591 Ev. 9704
-20 usec
20 usec
FA
ST O
utlin
e:
Why
do
you
want
todo
it?
What
isit
really?
How
do
you
want
do
toit?
When
do
you
pla
nto
do
it?
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
2)
FAST
Why?
Because GF is one of the three free parameters of the bosonicsector of the Standard Model. They are constrained by:
α(µ2 = 0) = 1/(137.035990 ± 0.000006) → (0.045 ppm)MZ = 91.1867 ± 0.0021 GeV → (23 ppm)GF = 1.16639 ± 0.00001 × 10−5 GeV−2 → (9 ppm)
† Note: the three parameters fully describe the strength of theelectroweak fields: γ, W and Z
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 3)
FAST
Parameter n. 1: α
- Josephson effect- Rydberg constant and de Broglie
wavelength of neutrons- Quantum hall effect- Anomalous magnetic moment of the
electron
(α-1-137.0360)x104
-0.40 -0.10 0.20
CODATA 1986 -.105 ± 0.061
h/Mn 0.113 ± 0.052
acJ -.230 ± 0.077
q-Hall 0.037 ± 0.027
ae -.001 ± 0.005
-0.40 -0.10 0.20
....but at nowadays energies the vacuumpolarization effects degrade this accuracy:
� � ��� � �� �
� ���
�
� � � �
� � � �
� � � � � ��� � � � �� � � � �� � � �� � ��� � � � � �
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 4)
FA
ST P
ara
mete
rn.
1:
α
Had
ronic
vacu
um
pol
ariz
atio
nm
easu
rem
ents
:
are
use
dto
extr
apol
ate
αat
hig
her
scal
es:
α−
1(M
2 Z)
=12
8.97
8±
0.02
7→
(209
ppm
)
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
5)
FA
ST P
ara
mete
rn.
2:
MZ
Nee
dle
ssto
say: 19
90-9
2
1993
1994
1995
√s [G
eV]
σ [nb]
L3
e+e− →
had
rons
(γ)
ratio
10203040
8890
9294
0.991
1.01
MZ
=91
.186
7±
0.00
21G
eV→
(23
ppm
)
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
6)
FA
ST P
ara
mete
rn.
3:
GF
Just
alitt
lere
min
d:
GF
Ã2
W+
g
g
ν e
µ+
ν µ
e+
µ+
ν µ
e+
ν e τ−
1µ
=G
2 FM
5 µ
192π
3(1
+∆
q)
∆q
incl
ude
the
hig
her
order
QE
Dan
dQ
CD
corr
ecti
onknow
nup
totw
o-lo
ople
vel(0
.5ppm
)
→T
.va
nRitbe
rgen
and
R.G
.Stu
art
Phys
.Let
t.B
437
(1998)
201
Unlike
α,G
Fdoes
n’t
suffer
had
ronic
unce
rtai
nti
es
whic
har
esu
ppre
ssed
by
m2 f/M
2 W:
mea
suri
ng
the
muon
life
tim
eis
aso
rtof
hig
h
ener
gyphysi
csex
per
imen
t!
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
7)
FA
ST P
ara
mete
rn.
3:
GF
Goi
ng
from
Fer
mim
odel
(con
tact
inte
ract
ion)
toSta
ndar
dM
odel
:
GF
=
√2g
2
8M2 W
(1+
∆r)
∆r
are
the
Ele
ctro
-Wea
kco
rrec
tion
s:
∆r
=f(∆
α,m
t,m
H)
=∆
α−
cot2
θ W∆
ρ+
...
Wpro
pag
ator
effec
ts(∝
O(m
2 µ/m
2 W)
∼0.
52ppm
)ar
etr
adit
ional
lyin
cluded
inth
edefi
nit
ion
ofG
F
Oth
erco
ntr
ibuti
ons
(∝m
2 tan
dlo
gm
H)
are
use
dto
der
ive
indir
ect
lim
its
onunm
easu
red
par
amet
ers.
The
pre
sent
valu
eof
δGF/G
Faff
ect
the
pre
dic
tion
for
MH
atth
eper
cent
leve
l.
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
8)
FA
ST P
ara
mete
rn.
3:
GF
Wher
eth
eunce
rtai
nty
onG
Fco
mes
from
?
δGF
GF
=−
5 2
δmµ
mµ−
1 2
δτµ
τ µ+
th.(
+4m
2 ν
m2 µ
)
5 2
δmµ
mµ
=0.
38ppm
PD
G2000
theo
ry=
0.50
ppm
Ritbe
rgen
and
Stu
art
4m
2 ν
m2 µ
=10
ppm
PD
G2000
(ifyo
uas
sum
eneu
trin
osto
be
mas
sive
)
δτµ
τ µ=
18ppm
PD
G20
00
Dom
inan
tco
ntr
ibution
sfrom
:Bal
andi
net
al.
(197
4)Bar
din
etal
.(1
984)
Gio
vanet
tiet
al.
(198
4)
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
9)
FAST
What is it really?
An experiment aiming to measure the muonlifetime with an accuracy of 2 ps, i.e., GF to 1 ppm
including all errors.
Requirements:
Large data sample of 1012 events over at least 9 τµ periods
⇓1012 × 9τµ = 1.5 years of data taking at 50 % efficiency if only one
event per cycle is accepted: parallelization needed!
⇓Tracking capabilities in order to disentangle overlapping events
⇓In order to stay in a reasonable data taking period ( ∼2 months), a
high data rate is needed: 1 MHz.
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 10)
FA
ST W
hat
isit
really?
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
11)
FAST
Systematics: learn from the past
Saclay experiment (Bardin et al.):A scintillator telescope surrounding a sulphur target in a pulsed pionbeam.
• The polarized µ component of the beam introduced time dependentinefficiencies due to the limited coverage of the detector
• The low granularity of the detector, together with the beam timestructure, produced time dependent effects
• Pile-up free background
TRIUMF experiment (Giovanetti et al.):
A water C̆erenkov system coupled to two PM in a DC muon beam.
• Electronic pile-up suppression: low statistics.
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 12)
FAST
Systematics: how to afford them
FAST wanted features:
• Capability to recognize the π → µ signature avoiding the“polarization effect”
• Controlled pile-up thanks to a high granularity
• Full solid angle coverage
• High detection efficiency over the complete positron energyspectrum
• Cross checked time measurements and stability
• Something else ?????
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 13)
FA
ST D
esi
gn
chara
cteri
stic
s
Beam
:
✔πM
1at
PSI:
DC
curr
ent
✔M
omen
tum
:17
0M
eV/c
(±3
%)
✔Siz
e:10×
16cm
2
✔In
tensi
ty:
106
s−1
✔P
uri
ty:
90%
Targ
et:
✔4*
4*16
00m
m3
scin
tillat
orbag
uet
tes
form
ing
the
read
out
pix
els
✔2
wav
eshifte
rsfiber
sper
bag
uet
te✔
25p.e
./pix
elfo
rm
inim
um
ionis
ing
par
ticl
e
✔40
×40
pix
else
nsi
tive
area
Readout
syst
em
:
✔10
0pos
itio
nse
nsi
tive
PM
wit
h4×
4phot
oca
tode
pix
els
✔ri
seti
me
<2
ns
✔1.
6K
dea
dti
me-
less
TD
Cch
annel
s
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
14)
FA
ST C
once
pt:
Asc
inti
llat
ing
pix
eldet
ecto
r,on
aD
Cpio
nbea
m,
able
tohan
dle
up
to30
π→
µ→
edec
aych
ains
ina
30µs
win
dow
:
plas
tic S
CIF
I ac
tive
targ
et
µ→eν
ν
µ→eν
ν
π→µν
π→µν
π+
π+
µ+ µ+
e+
e+
16 cm 10 cm
read
out p
ixel
(P
SP
M, 4
x4 m
m2 )
170
MeV
/c π
+
(1 M
Hz)
170
MeV
/c π
+
(1 M
Hz)
a) y
z vi
ew (
elev
atio
n)
b) x
z vi
ew (
plan
)
20 c
m
20 cm 20 cm
to P
SP
M v
ia
clea
r lig
ht-g
uide
s
beam
deg
rade
r (w
edge
)
beam
deg
rade
r (w
edge
)
scal
e
010
cm
z (b
eam
dire
ctio
n)x
(hor
izon
tal)
y (v
ertic
al)
z (b
eam
dire
ctio
n)y
(ver
tical
)
x (h
oriz
onta
l)
FAS
T
beam
co
unte
r
beam
co
unte
r
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
15)
FAST
Concept: from beam to decay time
☞ The pion stopping point is modulated by the plastic degrader suchthat the occupancy is kept uniform
☞ The pulse height in the stopping pixel is ∼ 5 times higher than am.i.p. ⇒ a double threshold discriminator (developed at PSI)is used for tagging purposes
☞ Every event is a good event and the readout system is almostdead-time free. So, why do we need a trigger?
✧ t0 for the whole system;✧ definition of a ”region of interest” where to look for decaying
muons and decrease the data rate.
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 16)
FA
ST Tri
gger
syst
em
1st
level
:th
ree-
fold
coin
ciden
cedefi
nin
gth
et0
2n
dle
vel
:tw
osn
apsh
ots
ofth
eev
ent
(15
ns
and
100
ns)
are
use
dto
find,re
spec
tivel
y,th
est
oppin
gpoin
tofth
epio
nand
muon.
Only
the
5×
5su
per
pix
elaro
und
the
stoppin
gpoin
tw
illbe
analy
sed.
For
each
row
a100
MH
zco
ntr
oller
searc
hfo
r“st
oppin
gpatt
erns”
:
stop
and
dec
ayπ
µde
cay
20
s µ
0 (t
rigg
er)
15 n
s1s
t
5 ro
w tr
igge
r ro
ads
rese
t
100
ns2n
d tr
igge
r sn
apsh
ot
y sy sy 0
z s
z s
colu
mn
OR
a)
b)
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
17)
FA
ST R
eadout
syst
em
:P
SP
M
The
sign
alis
tran
spor
ted
from
the
scin
tillat
orbag
uet
teto
the
100
Ham
amat
suPos
itio
nSen
siti
veP
M’s
via
two
gree
nw
aves
hifte
rfiber
s:
4x4
x1
60
mm
3 b
ag
ue
tte
(Bic
ron
BC
40
0)
BC
60
0 o
ptica
l ce
me
nt
BC
62
0 w
hite
diffu
se
refle
ctive
pa
int
1m
m d
iam
ete
r B
icro
n 6
92
(fa
st)
gre
en
wa
ve
sh
ifte
r fib
re
Eac
hP
M,pac
kage
dw
ith
HV
div
ider
and
pla
stic
cove
r,pro
vid
es4×
4=
16an
odes
(chan
nel
s),
4×
4m
m2
each
:
17 m
m (
17 fi
bres
)
17.5 x 17.5 mm2
30 x
30
mm
2
17.3 mm (20 fibres)
π+ dire
ctio
n
PS
PM
en
velo
pe
pixe
l (an
ode)
4 x
4 m
m2
bea
m
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
18)
FAST
Readout system: TDC
The stopping point coordinates are used to mask the hits digitizationin the heart of the system:
✔ 64 TDC chips developed by CERN/ECP-MIC group andengineered by CAEN on 16 128-channel boards (v767)
✔ “0” conversion time
✔ up to 520 ps of time resolution
✔ up to 1 MHz rate capability
The TDC system is driven by an external Rubidium clock with a baseoutput frequency of 30 MHz and a stability of ∆f/f = 2.2 × 10−10,
much beyond the requirements.
A second Rubidium clock is used as an external time reference andcalibration system.
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 19)
FAST
DAQ system: a challenge
The Data AcQuisition system must sustain a continuous event rate of 1 MHz.
From physics to tape:
☞ 60 hits/event + control words → 420 Mbytes/s☞ using LV2 info (+ tricks) → 80 Mbytes/s
Note: only a 5 × 5 superpixel region is used. We are forced to drop the rest.
An unfair comparison:
LHC detector FAST
n. channels 100,000,000 2304event size 1 MByte 500 BytesPhysics data throughput 100 MBytes/s 420 MBytes/s
The data are sent to a PC farm which perform the analysis and stores histograms.Only a fraction of the event is recorded for systematics checks.
The full (LV2 info) saving would require a huge amount of mass storage(∼ 80 Tbytes) filled with W.O.R.N.’s†.
The reading+processing time would require much longer than a newdata taking period.
†W.O.R.N.: Write Once Read Never
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 20)
FA
ST D
AQ
syst
em
:a
challenge
Aco
mm
erci
ally
available
CP
U-les
sso
luti
on
has
bee
nadopte
dfo
rth
eV
ME
syst
em:
✦P
VIC
=P
CI
toV
ME
Inte
rcra
teC
onnec
tion:
ades
kto
pP
Cis
use
das
are
mote
CP
Ufo
racc
essi
ng
VM
Eaddre
sses
✦H
igh
PC
Idata
rate
→80
MB
yte
s/s
✦H
igh
VM
Edata
rate
(tes
ted
inla
b.)
→18M
Byte
s/s
✦4
VM
Ecr
ate
s+
4co
ntr
oller
PC
’sare
suffi
cien
t
T D C
V M E P C I
T D C
T D C
T D C
T D C
VM
E b
us
T D C
V M E P C I
T D C
T D C
T D C
T D C
VM
E b
us
T D C
V M E P C I
T D C
T D C
T D C
T D C
VM
E b
us
T D C
V M E P C I
T D C
T D C
T D C
T D C
VM
E b
us
PC
PC
PC
Con
trol
−PC
Ana
lysi
sF
arm
Ana
lysi
sF
arm
Ana
lysi
sF
arm
PC
Ana
lysi
sF
arm
PV
IC b
us
Fast
−E
ther
net
Fast−Ethernet
direct connection
LucaM
alg
eri
FA
ST
:an
experim
entfo
ra
hig
hpre
cisio
nm
uon
life
tim
e(a
nd
GF
)m
easu
rem
ent
(page
21)
FAST
How do you want do to it?
Simulations have been performed to study efficiencies and systematics:
1) Trigger configuration and efficiency51 % of the incoming pions are triggered within the two snapshotstructure. The losses are mainly due to early decaying pions, beforethe second snapshot takes place. Muons are identified by means ofthe pulse height signal.
2) Selection algorithm85 % of the events have one or more particles originating fromother decay chains and traversing the same 5 × 5 pixels region. Inorder to reduce this accidental background component, positrontopology cuts have to be applied
=⇒
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 22)
FAST
How: selection and efficiency
✦ Hits in the 5×5 pixels region areclustered in time
✦ A positron track is defined as atrack with hits in both the inner(3× 3) and outer (5× 5) region
✦ Only one cluster of less than 4pixels is required in the outerlayer
✦ Events with more than onepositron track are rejected
Total efficiency: 25 %Accidentals: 3 × 10−4
– Total a) all events– Muons b) positrons tracks– Accidentals c) only one positron track
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 23)
FAST
How: systematics - muon spin rotation
A polarized muon source (from beam or from asymmetric muondetection efficiency) precesses around a magnetic field (Earth). IfALSO the positron detection efficiency is asymmetric this can cause atime dependent effect:
µ spin direction
spin t=0
spin
t=t1
positron
positron
B field
cos θ (e+)
cos
θ (e
+)
even
ts
φ (e+)
φ (e+)
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 24)
FAST
How: systematics - muon spin rotation
✔ Avoid polarized beam muons requiring a detectable π → µtransition.
✔ Make the detector as uniform as possible:
- residual muon detection anisotropy along the beam direction fromdouble hit time resolution and high threshold discriminator (10 %)
- positron detection inefficiency from local effects (unstablethresholds/gain → 1 %)
Precession frequencies:
νµ = 13.55KHz
G× B νmuonium = 1355
KHz
G× B
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 25)
FAST
How: systematics - muon spin rotation
Residual of time distributions after applying a weak
magnetic field (50 G):
Time [µs]
Res
pons
e (p
pm) a) B field parallel to fibres
- generator
a) B field parallel to fibres - experiment
b) B field perpendicular to fibres - generator
b) B field perpendicular to fibres - experiment
0
0 2 4 6 8 10
Time [µs]0 2 4 6 8 10
Time [µs]0 2 4 6 8 10
Time [µs]0 2 4 6 8 10
255075
100
-100-75-50-25
Res
pons
e (p
pm)
0255075
100
-100-75-50-25
Eve
nts
(ppm
)
0255075
100
-100-75-50-25
Eve
nts
(ppm
)
0255075
100
-100-75-50-25
With a suitable choice of the magnetic field and a proper fit procedurethis effect can be kept at the level of (0.2 ppm).
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 26)
FAST
How: systematics - time dependent detection efficiencies
✦ After-pulses, in a 10 ns scale, could fake a muon pulse and affectthe requirement of a fully detected π → µ → e chain.
Prob. → 10−4 with double threshold discriminator
✦ On a longer time scale (0.5 µs), after-pulses, gain changes andbaseline shifts may affect the electron detection efficiency in pixelsalready visited by a pion or muon.
Low Threshold probability
≥ 1 pe 10 %≥ 2 pe 1 %≥ 3 pe 0.1 %
The decay time distribution gets two additional terms:
1) ∝ exp(
−t 2
τµ
)
: prob. for an electron to be just under threshold
2) ∝ exp[
−t(
1
τµ+ 1
τβ
)]
: τβ is the slow after-pulse time component
Simulations based on lab. tests give a total effect of less than 0.3 ppm.
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 27)
FAST
How: systematics - tracks overlap and pile-up
Given the high occupancy of the detector, event and single trackoverlap have to be taken into account.
Two separate decay chains (independent on each other):
(π1, µ1, e1) ⇐⇒ (π2, µ2, e2)
may “interact” and give rise to fake decay chains:
(π1, µ1, π2), (π2, µ1, e1), (π1, µ2, π1),. . .
Most of these backgrounds have either flat or well behaved
∝ exp(
− tτµ
)
time distributions.
→ no or little effect on final measurement
but for some of them .......
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 28)
FAST
How: systematics - tracks overlap and pile-up
(π1, µ1, π2): the positron e1 is lost and an incoming π2 fakes it. If the following decayappear in the predefined time window [−10 µs; 20 µs] the event is not accepted(→ multiple-track): late pions have better chance to be accepted
(π1, µ1, e2): mirror component, with the positron coming from a (previous) very late decayin the same super-pixel
Decay time [ns]
t max
τµ
− t− )-(exp
τµ
t minexp(- )+ t
Flat acc.
Using tmin = −10 µs, tmax = 20 µs, and the probability forthe pion track to be reconstructed as a positron track(Montecarlo), this effect is found to be negligible andunder control studying the negative time distribution region.
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 29)
FAST
How: systematics - tracks overlap and pile-up
A little bit more complicated case: double kill
The presence of two events with same (or somehow related) space and timestructure may give rise to time dependent inefficiencies:
prob(ekill) ∝ 1 + af(te2 − te1) prob(πkill) ∝ 1 + a′f ′(tπ2− tπ1
)
eff. = 1 − af(te2 − te1) − a′f ′(tπ2− tπ1
) + aa′f(te2 − te1)f′(tπ2
− tπ1)
The first two “single kill” terms are NOT dependent on (te1 − tπ1), while the third
gives an additional contribution ∝ exp (−t/τµ).
π2 ↔ π1 correlations:
- π2 enter the detector just after π1 during the trigger dead-time
- π2 enter the detector just after π1 and TDC cannot resolve the two hits
- π2 enter much before π1 when pile-up rejection doesn’t apply (< −10 µs)
e2 ↔ e1 correlations:
- e2 overlaps in time and space e1
under control using the beam rate → 0.2 ppm
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 30)
FAST
How: systematics - physics effects
Muonium formation in scintillator is highly probable (66 %). Is themuon lifetime bound in Muonium state different?
✔ Early studies claimed YES
✔ A more recent paper from Czarneki, Lepage and Marciano(hep/ph-9908439) show that the correction is at the level of 10−9,in agreement with phase space arguments
✔ Negligible effect on FAST measurement
Rare decays:
✔ Radiative decays may alter the topology of the event and hence thedetection efficiency
✔ Full GEANT simulation show negligible effects
✔ Additional electrons from γ conversions increase the number ofpositron tracks in the detector by 0.25 % → negligible effect
✔ Internal conversions µ+ → e+ν̄µνee+e− produce an efficiency loss of
< 3.4 × 10−5 but same time structure as the signal
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 31)
FAST
How: fit and summary
Including all considered systematic effects the final fit function willlook like:
N(t) =N0
τµ
δt
[
P0 + exp
(
− t
τµ
)
+ P1 exp
(
− 2t
τµ
)
+ P2 exp
(
+t
τµ
)]
⊗ µSR
P0 =flat accidentalP1 =double killP2 =late pions
P0 and P2 can be left free in the fit but P1 wouldincrease the systematic error by 2-3 ppm.Need to get it from data: dedicated runs for effi-ciency studies, beam rate, off-line extrapolation.
Systematics summary tables
Source Syst. error (ppm)Before After corr.
Muon spin rotation 1.7 0.2Time dep. efficien. 0.9 0.3Track overlap 1.1 0.2Physics effects - 0.001
Total systematics 0.41
Source Error (ppm)
Statist. + acc. 1.2Systematics 0.4
Error on τµ 1.3
Radiative corr. 0.2
Error on GF 0.66
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 32)
FAST
When do you plan to do it
First milestone met at the end of year 1998: test-beam at PSI
✦ Check feasibility studies
✦ 1/10 of the detector with several configurations
✦ Low intensity
✦ Check fibers, PSPM and TDC’s behaviors
Run 7591 Ev. 9704
-20 usec
20 usec
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 33)
FAST
Test beam
The dynamic range has been studied, together with double hitresolution, discriminator requirements, TDC performances, etc.
0
2000
4000
6000
8000
0 250 500 750 1000 1250 1500 1750 2000
Mean 170.7
Energy loss for positrons (arb. unit)
N/1
8
a)
0
100
200
300
0 250 500 750 1000 1250 1500 1750 2000
Mean 456.1
Energy loss for incoming pions(arb. unit)
N/1
6
b)
0
200
400
600
800
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Mean 948.5
Energy loss for stopping pions(arb. unit)
N/2
0
c)
193.4 / 191
P1 8.738 0.4386E-02P2 -0.4543E-03 0.1320E-05
τµ = (2201.0 +- 6.4 ) ns
χ2/d.o.f. = 1.01
τµ(PDG) = (2197.03 +- 0.04) ns
10-1
1
10
102
103
104
0 4000 8000 12000 16000 20000Time (ns)
No.
of µ
→e
deca
ys
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 34)
FAST
When do you plan to do it: cont’d
Date Activity
✔ 2000 (1st) Experiment approved at PSI✔ 2001 (1st) Prototype tests, re-design of DAQ
2002 (3rd) System integration tests at PSI2003 (2nd) Installation2003 (3rd) Check-out2003-2004 Data taking
Collaboration: CERN, University of Geneva, NIKHEF, PSI
http://www.cern.ch/fast/
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 35)
FAST
Conclusions
✦ FAST aims to measure with 10 times better accuracy than thepresent world average the Fermi coupling constant GF
✦ Major challenges:
- stability and data rate: from a 0.008 m3 detector, an LHCdetector equivalent throughput has to be handled → a test benchfor new generation experiments
- very subtle systematic effects have to be sorted out
✦ Feasibility established in simulation and test beams
✦ Stay tuned for results in 1-2 years
Credits for the material:
A. Barczyk, F. Cavallo, P. de Jong, P. Kammel, J. Kirkby, R. Nahnhauer, F.
Navarria, G. Passaleva, A. Perrotta, C. Petitjean, M. Pohl, R. Stuart, G. Valenti,
D. della Volpe
LucaMalgeri
FAST:an experiment for a high
precision muon lifetime (and GF )measurement
(page 36)
Recommended