Transcript ppt
Some ACD Calibration issues
I.
ACD mip peak monitoring
1.
ACD mip peak monitoring requires the measurements to be done at
similar conditions. The best is to use normal incidence (to the ACD
tile plane) single charged particles (mip)
2. In ACD standalone tests (TriggerOps):
a) up-right position (Z up); quasi-normal incidence muons
events are selected by coincidence of
signals from any two pair of sides (top
is also considered as a side). Main test,
used for mip peak position monitoring
b) horizontal position (X or Y up). Coincidence between
upper and lower sides. Used for light yield
measurements for the tiles on upper and
lower sides
These tests were performed at Goddard during environmental and
pre-shipping tests, and at SLAC.
Alex Moiseev NASA/GSFC 12/16/2005
Now we have to find a way to compare mip peak positions
obtained in ACD standalone tests with that obtained with upright LAT triggered by L1T:
Select events by the incidence angle, or by the reconstructed
trajectory pathlength in the tile
AcdTkrIntSecPathLengthInTile (SVAC ntuple variable)
Do it for up-right ACD (Z up) with L1T and TriggerOps (need
TrigOps!) Doing this we will make connection of all previous
ACD standalone tests with that in LAT
Use obtained mip peak positions with L1T trigger as new
reference values
The same event selection shall be done with LAT in orbit
Alex Moiseev NASA/GSFC 12/16/2005
What we will have with L1T?
Top Tile
All events
(3 % of total statistics)
Selected by the path in the tile
9.9mm < Path < 12mm
Survive ~ 30% events
Side Tile
All events
(1.7% of total statistics)
Selected by the path in the tile
9.9mm < Path < 12mm
Survive ~ 3.5% events !
See how mip peak moving with
applying path length selections!
To collect ~2000 events in side tile (needed for reliable mip peak fitting) we
need to collect ~3.5 M events
total,NASA/GSFC
which should
Alexin
Moiseev
12/16/2005take ~2 hours with the
rate of 500Hz. Looks reasonable
II. Edge effect measurement
Light yield from the tile decreases in the tile edge area. We run
bench tests at Goddard and estimated that the light yield at the tile edge
is ~70% of that in the tile center, and recovers to 100% at 3 cm from the
center. But that measurements were rather crude, using scintillating
fiber hodoscope to map the light yield edge effect.
Here are MC simulation results on edge effect (based on bench tests
results). Tests with LAT will provide us more accurate data.
PHA vs.
AcdActiveDist
Edge effect
PHA vs.
AcdTkrIntSecLocalX(Y)
Edge effect
Alex Moiseev NASA/GSFC 12/16/2005
III. ACD tile efficiency measurement.
Knowledge of tile efficiency is needed to estimate the absolute
light yield (in photoelectrons), which is to be placed in ACD
simulation to determine overall ACD efficiency.
This is how it was done in ACD Performance tests at Goddard:
Triggering
hodoscope
Tested TDA
ACD
Incident
muon
Alex Moiseev NASA/GSFC 12/16/2005
We want to do the same measurements with LAT to confirm
our light yield knowledge.
This is what I obtained from surface muon MC run:
Top Tile:
Selections applied:
–
–
–
–
TkrNumTracks=1
9.9<AcdTkrIntSecPathLengthInTile<12
AcdActiveDist>40
AcdTkrIntSecTileId=current tile
Seems that clean normal incidence
tracks hitting the central part of the tile
should pass these selections.
But I cannot remove these events!
I tried to select events with only one track, checked ChiSq, checked
number of tracker hits – does not help.
I cannot determine tile efficiency due to presence of these events!
Alex Moiseev NASA/GSFC 12/16/2005