[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
BCAL Dark Hits
Hi all,
I've just committed code to the repository that generates dark hits on
the BCAL sensors.
Briefly, the algorithm is:
For each of sensor in the inner layers only ( N_sensors = 2 * N_cells ):
- determine the number of dark pulses from random poisson with mean =
( dark rate ) * ( ADC integration window )
- generate additional cross talk pulses from random poisson with mean
= ( number from above ) * ( cross talk probability )
- convert sum of both of these to an energy based on the photoelectron
< - > energy conversion
If either of the sensors of a cell already has a hit, then add this
extra energy to the hit. Otherwise, create a new hit with this energy
and a time that is randomly picked uniformly between -25% and +75% of
the ADC integration window.
A few notes:
The relevant parameter to adjust the dark rate is
BCALRESPONSE:DARK_RATE_GHZ. The nominal value is 0.041 (41 Mhz), i.e.
to get back to the rosy response do -PBCALRESPONSE:DARK_RATE_GHZ=0.
Hits for upstream and downstream sensors are handled independently.
The reconstruction code requires a hit in both ends in order for a
cell to be used in reconstruction -- this requirement provides some
suppression for spurious cells since one must have a dark pulse above
threshold in both the upstream and downstream readouts.
The algorithm for determining the threshold depends on dark rate (see
the GlueX-doc 1231). This means that as the dark rate is turned up,
the relative number of cells with spurious hits should stay constant
at the fraction set by: BCALRESPONSE:OCCUPANCY_FRACTION_LIMIT.
However, the energy spectrum of these spurious hits (because the now
contain more dark pulses) is higher.
Attached is a plot of "hybrid" pi0s -- one photon in BCAL and one in
FCAL taken from a small sample of 10k PYTHIA events. There are no
cuts on this sample. You can see what the peak looks like for three
different dark rates. Clearly this is only the tip of the iceberg.
There is energy dependence to study as well as how can the
reconstruction be tuned (minimum cell energy cut) etc. to accommodate
various dark rates. My goal was to get some working code into the
repository so we can begin to explore this.
-Matt
dark_rate_compare.pdf