HF test beam results

1
HF test beam results

• Two wedges assembly (H2 beam)

2
Test beam setup
Iron shield
Two Hf wedges Four data taking conditions-Perpen
dicular to beam-Inclined by 6o
horizontally-Inclined by 4o vertically-Modules
turned by 90 degrees to scan along the side
Muon tagger
Delay Wire Chambers Resolution 200 mm
Trigger counters
3
General layout
Linux Midas
NIM
ROBOX
Lecroy 2774 preampl.
CAMAC
F I P
F B D
FASTBUS
1 8 8 5
Ethernet
Spare space
4
Test beam continued
The pulse from the PMTs operated at a voltage
corresponding to a gain of 105- was fed trough a
LRS 2094 amplifier card (gain 60) HF dedicated
Data Acquisition FASTBUS based with Pulseheight
read through FB ADC with 50 fC/count sensitivity
and DWC readout through TDC with 1 ns resolution.
The acq setup was completely independent from
the one of the HB test which has allowed optimal
usage of the test beam time. Acquisition of
source electrometer via special CERN built CAMAC
ADC (courtesy of SL/BI group) Data collected
with electrons and positrons with energy between
10 and 250 GeV and with pions between 10 and 300
GeV. Large samples of muons have also been
collected. The wedges were installed on a
movable table computer controlled. Typical rates
were lt10000 particles/spill with gt300 triggers
logged /spill
5
Test beam layout
Seen from above
Beam 0 and 6 degrees
beam
Seen from side
4 degrees
6
How much light ?
10 GeV e- 1 p.e. 90 ADC counts ltp.e.gt10
GeV227/902.5 The 0 p.e. peak (8.6 of events)
matches what one expects from a poissonian
distribution with average 2.5 ! The pedestal
width (lt3 counts, i.e. lt30 MeV equivalent) is
negligible with respect to the single p.e. signal.
Ph (ADC counts)
The light seen in the electromagnetic section of
HF is 0.25 p.e./GeV
7
Light collection problems

• One of the issues was the determination of the
  light yield of HF with the final sampling
  fraction. Shortly after deploying the wedges onto
  the beam two problems were discovered
• For some towers the reflector in the air-core
  light guide had been installed on the wrong less
  reflective- side
• The geometry of the ROBOXes did not match the
  length of the light guides leaving a gap of 17mm
  between the end of the light guides and the
  photocathode. This problem reduces the amount of
  light by almost a factor of 2.
• Due to the time pressure these problems were
  fixed only for one channel (tower 15) which is
  the one used to estimate energy resolution etc.
  The reflector is on the right side and the gap is
  2 mm. When looking at results on other
  channels one has to keep in mind that the light
  seen is off by a factor of 2

8
e- and p- see the same calorimeter
Tower 15 EM
10 GeV e-
10 GeV p-
Ph (ADC counts)
9
HF as seen by m
ltPH gt(adc counts)
Middle of tower
m see fiber structure !
X impact (cm)
Evgenii Vlassov, CERN
10
m continued
Max 40 loss Over 2-3 cm
The crack between wedges as seen by muons
Evgenii Vlassov, CERN
11
Electron energy resolution
Clear problem seen with 30 GeV electron run
junked
Evgenii Vlassov, CERN
12
Electron energy resolution
a 2.09, b0.06
a 2.06, b0.01
Tower 15 , Wedges inclined vertically by 4
degrees
Evgenii Vlassov, CERN
At 100 GeV resolution has an additional 8
fluctuation (added in quadrature) wrt what is
expected by p.e. statistics
13
Electron energy resolution
Beware HF will have features in resolution
varying with rapidity see below the fiber
structure very visible when the beam is
perpendicular to the wedge ( what will happen at
high rapidity at h5 the angle is 0.6 degrees)
14
Linearity
Evgenii Vlassov, CERN
p.e./GeV
Tower 15
Linearity OK (take into account that there is
also some uncertainty on the beam energy !)
/- 3
Energy (GeV)
15
Crack seen by e
Saeid Paktinat, Teheran
25 max loss over 2 cm Local structures non
always understandable work in progress
?
16
Hadronic energy resolution
Preliminary study Use only EM tower where beam
was centered Gaussian fit s/E levels at 20
17
The cracks as seen from p-
ltPHgt (adc counts)
Max 15 loss Over 3 cm
Evgenii Vlassov, CERN
Energy (GeV)
18
Source calibration
Source calibration successfully implemented in
parallel to test beam exposure both wedges in
H2 have been systematically sourced. A 3 mCurie
Co60 source is pushed into source tubes ( at
least one per tower) and current drawn by PMTs is
measured by electrometer. The results are
compared to the longitudinal scan done with the
100 Gev electron beam.
Under fiber bundles
Inside Backplane
Current ( arb units)
Under PMT
Inside absorber
Source pos.
19
Attenuation Studies Source
Redpushing
nA
Precision limited by sensitivity of electrometer
to noise Clear evidence of attenuation. The
average over the channels measured is

3.7 /- 1.2
Blueretracting
Distance (cm)
20
Attenuation studiesbeam
Evgenii Vlassov, CERN
Wedges turned by 90O wrt to beam and scanned
longitudinally with positron beam Pulse height
very different (shower start, shower max, shower
end) Conclusions Attenuation length lt 2/m more
cannot be said due to systematics of measurement
Both Source and Beam measurement seem much closer
to the att. Measured with light sources (3/m
_at_400 nm) than to what is measured with source at
TTU
P.H. Adc counts
21
Raw data online
Unpacked ntuples /shift/cmsc01/data6/HF/ntuples/
run.rz 10 Gb Packed ntuples /afs/cern.ch/cms
/HCAL/HF/ntuples/run.rz.bz2 2Gb General
information and some results http//cms-testbeam
h2.web.cern.ch/cms-testbeamh2/HF/ Short logbook
runs.txt (there is at all the 3 adresses) My
e-mail e.vlassov_at_cern.ch