At ILC, 2820 bunches of electron and positron beam
make collisions successively with 337ns bunch intervals.
This succession of 2820 bunches is called "train", and
trains are repeated at a rate of 5Hz.
Due to low energy
electron/positron beam background,
the hit rate of the innermost layer of the
vertex detector is estimated
to be ≈1.5 hit/cm2
at R=2.0 cm and B=3T for one
bunch crossing (BX). If the hits are accumulated
for one train, the hit density becomes very high, and
the pixel occupancy for a pixel detector with
25μm pixel size exceeds 10%, which is
One method to keep the pixel occupancy acceptable
level (≈0.5%) is to read out the sensors more than
20 times in one train.
There are R&D activities by several groups
based on this method. This method requires
very fast readout speed, and the feasibility
has not been demonstrated yet.
In the FPCCD vertex detector, we use very fine
(≈5μm) pixel CCDs. The pixel occupancy
will be less than 0.5% even if the hit signal
is accumulated during a total train of 2820 bunches.
In order to suppress the number of hit pixels
due to diffusion in the epitaxial layer,
the sensitive layer of the FPCCD should be
Another advantage of the FPCCD vertex detector is
that it is completely free from electro-magnetic
interference (EMI) caused by the beame.
At the SLD, an experiment executed at SLC which
is the first linear collider in the world,
the EMI caused by the beam affected the readout
of the vertex detector. For the FPCCD vertex detector,
however, the EMI would cause no effect because
the detector is read out between trains.
Recently, CCDs with the pixel size less than 5μm
are used for digital cameras. So, we believe the FPCCD
is technically the most feasible option for the vertex detector.