Download - Status of the TPG
Status of the TPG
E.RadicioniMICE coll. Meeting - RAL, 27-29 Oct.
2004
27/10/2004 E.Radicioni 2
Outline
• Recall of the main characteristics• Construction and operation• First of the detector behavior• Design aspects
– Electronics– Field-cage– Gas– Background
27/10/2004 E.Radicioni 3
• The TPG is a cylindrical TPC equipped with GEM amplification and a special high-resolution pad-plane
• TPCs with GEMs are actively studied by several groups around the world in view of the linear collider experiments.
• The novelty of the TPG consists in the pad-plane, where projective readout promises high granularity with a reduced number of channels
• Dedicated design choices for MICE were – the dimensions (100cm length, 30cm diameter) to fit
into the tracker magnet– The gas (He) chosen because of its low conversion
probability
27/10/2004 E.Radicioni 4
Construction
E
B
drift volume
field cage
GEM stage solenoid
read-out plane
27/10/2004 E.Radicioni 5
The GEM foil• Triple-layer structure• Each layer divided in
8 regions, independently powered
• Max amplification depends on gas and HV. With good gas it can be as high as 106.\
27/10/2004 E.Radicioni 6
The hexaboard
– ~710000 hexagonal pads
• size: 300 m • pitch: 500 m
– grouped into strips along 3 coordinates at 120 degrees (u, v, w) running at different depths
27/10/2004 E.Radicioni 7
Strips and pads
27/10/2004 E.Radicioni 8
300 m
500 m
27/10/2004 E.Radicioni 9
Spot the error … and correct it!
27/10/2004 E.Radicioni 10
Hexaboard quality, from this:Layer N
for shorcut: C==0
-20
0
20
40
60
80
100
120
140
0 100 200 300 400 500 600
strip number
C [
pF
]
n:n+1
connector
27/10/2004 E.Radicioni 11
… to this:
27/10/2004 E.Radicioni 12
Support structures
1
2 34
5
1. outer metal ring
2. o-rings for gas tightness
3. stiffener plate4. GEMs support
rings5. guard ring
27/10/2004 E.Radicioni 13
27/10/2004 E.Radicioni 14
Assembling the detector
27/10/2004 E.Radicioni 15
27/10/2004 E.Radicioni 16
27/10/2004 E.Radicioni 17
The dream team at work
27/10/2004 E.Radicioni 18
Last computations and timing checks …
27/10/2004 E.Radicioni 19
... what is this detector trying to tell us ???
27/10/2004 E.Radicioni 20
55Fe source calibration
• Absolute energy, equalization• Preamp gain calibration is not included
yet• First indications are
– Despite the lack of gain equalization, the energy resolution is already quite good
– The plots are very sensitive to the cuts on cluster size, as expected from a properly working detector
27/10/2004 E.Radicioni 21
55Fe source
• “beam profile”• Unconnected strips• Ill electronic
channels– Dead– Noisy
27/10/2004 E.Radicioni 22
• Capability of seeing detector structures
• 55Fe source illuminates the center of the pad plane
• Large lines: spacer grid in between the GEMs: 2mm large
• Thin lines: non-working electronics channels
27/10/2004 E.Radicioni 23
55Fe source
All clusters
≥2 strips
≥3 strips
27/10/2004 E.Radicioni 24
U,V vs. W correlations• out-of-line events are
due to lack of gain calibration
• Correlation is important: it can be exploited as an additional tool for getting rid of fake combinations– In addition to the use of
the 3rd projection
• Compass is able to reject (almost all) fakes by this technique
TPG
COMPASS
27/10/2004 E.Radicioni 25
Low energy tracks
• 500 samples @10MHz
• Ar/CO2 90/10, 10cm/s total 50cm drift path
• e- from Sr source• B=0.07T (1/10
nominal)• Color code gives
charge amplitude
27/10/2004 E.Radicioni 26
• 2 MeV/c electron in B=0.07T
• Transverse diffusion spreads the charge
27/10/2004 E.Radicioni 27
Intrinsic resolution• 55Fe X-ray conversion
position can be determined by 2 projections, then cross-checked with the 3rd one.
• The intrinsic resolution is VERY promising
• This has been obtained with a 3cm drift cell.
• Actual resolution over longer drift depends on gas properties.
~40µm
U+V vs. W resolution
27/10/2004 E.Radicioni 28
Design improvements
• Electronics• Ingredients for good resolution• Gas choice• Discussion on overall parameters• A compact emittometer, alias TPG ?
27/10/2004 E.Radicioni 29
Electronics
• New electronics from ALICE• Higher integration• Total (including DAQ) 10 CHF / channel• Digitization is close to detector (less
noise)• Data get out from the detector on a few
optical fibers (elegant, simple)• Large range of possible sampling
frequencies up to 40 MHz.
3027 February 2004 Luciano Musa / CERN - PH
ALTRO EVOLUTION
4cards16 ch
13
5 m
m
1998
CHANNELS / CHIP: 1
POWER / CH: 120mW
PRICE / CH: 50CHF
Integrated ADCs2
0 m
m
4 PQFP 100
8 SSOP 28
24 mm
1999
CHANNELS / CHIP: 4
POWER / CH: 80mW
PRICE / CH: 8CHF
2001
CHANNELS / CHIP: 1
POWER / CH: 16mW
PRICE / CH: 5CHF
3127 February 2004 Luciano Musa / CERN - PH
Digital Conditioning of the TPC Signal
EV 1 EV 2
EV 3
EV 1 EV 2
EV 3
EV 1 EV 2EV 3
EV 1 EV 2
EV 3
ADC BC I
TCF BC II
27/10/2004 E.Radicioni 32
Momentum resolution• Position resolution is driven by
– Readout pitch (fixed)– Diffusion in gas– Ionization and gain
• Gas choice is a key point: we should look for larger gain and ionization and smaller diffusions.
• Momentum resolution is driven by– Position resolution– Number of available points
• From the simple Glukstern formula: 50 points give dp/p resolution factor only 1.5 better than 20 points.
• If more distant points are affected by more diffusion, we should really consider limiting the number of points and the drift length.
27/10/2004 E.Radicioni 33
Gas choice
27/10/2004 E.Radicioni 34
Shorter field-cage?He/CO2 1m Ne/CO2 18cm
E 500 V 300 V
Max HV 50 KV 5.4 KV
Drift time 60 s 6 s
Drift velocity 1.68 cm/m 3 cm/m
Sampling freq 2MHz 10MHz
Number of samples 118 60
Trans/long diffusion 1000/1600 m 80/200 m
Specific ionization 10 e-/cm 20 e-/cm
Usable long. Slices 118 20 (shaper limited)
N. Radiation lenghts 6.6 E-4 5 E-4
X-ray abs. Coeff. 2.5 E-5 cm-1 1.2 E-4 cm-1
X-ray abs. probability 1 0.4
Electronics HARP ALICE
27/10/2004 E.Radicioni 35
Tracks and Background in the short TPG
• Intrinsic occupancy limit of the detector: N.strips/cluster size ~ 200– By exploiting correlations (a-la-compass) one could
maybe do better
• 1 muon per s 6 tracks in one “image”• Number of background tracks (normalized per
MHz imput photons) ~ 0.013 MHz-1
• Quoting Rikard’s computations: 22MHz of photons entering the tracker volume in the upstream spectrometer if this rate is OK there will not be any trouble.
• Photon conversion in the HV membrane could account to the same number as conversions in gas not much of a difference.
27/10/2004 E.Radicioni 36
Conclusions• The TPG detector is operational in the HARP test-
bed• Placed an order for 2000 channels of the new
electronics• First results indicate performances at the
expected level or better• Several design parameters can be reviewed
– Better gas choice better resolution– Shorter detector less material, straightforward
construction, compact, could stay in shorter solenoid.– Faster drift time less X-ray background– Better resolution
• More results from beam tests in the near future