Today's Progress 27. June. 2006

Tune of PB time walk by (K^-,X) fixing PA time(1)

Method

Since PA time walk is fixed, and T0-PA TOF gate is now available for 100% of the production run, we try to follow the PB time walk by delayed-event, defined by the condition,

deltaT'(T0->PA) = Tpa' - Tt0' -TOFkstop -TOFsec > 0.8 (nsec) ,

after first stage PA/T0 time walk correction on Tpa/Tt0, i.e.

Tpa' = Tpa - PAOFFSET(idrun)

and

Tt0' = Tt0 + T0OFFSET(idrun).

The arm-by-arm behavior of the Kmu2 peak under the event selection above is exhibitted below. Note that PA time walk is not corrected for PA-PB TOF below.

time walk of T0 2nd layer

Now, we examine the run-by-run behavior of the Kmu2 peak taking the PA time walk determined by T0-PA TOF analysis into account, namely,

delta T (PA->PB) = Tpb - Tpa -TOFkmu2

->

delta T' (PA->PB) = Tpb - Tpa'-TOFkmu2

Note that we consciously keep old 1/beta to calculate deltaT (T0->PA).

Now, delta T(PA->PB)/delta T' (PA->PB) center and width are exhibitted and compared.

The run-by-run variation of the gaussian center of delta T (PA->PB) for L(black) and R(red) arms.
time walk of T0 2nd layer
The run-by-run variation of the gaussian width of delta T (PA->PB) for L(black) and R(red) arms.
time walk of T0 2nd layer
The run-by-run variation of the gaussian center of delta T' (PA->PB) for L(black) and R(red) arms.
time walk of T0 2nd layer
The run-by-run variation of the gaussian width of delta T' (PA->PB) for L(black) and R(red) arms. A substantial improvement is seen for Rarm after run 210, where sudden jump of PA timing has been found by T0-PA analysis.
time walk of T0 2nd layer

We activate the PB run-by-run, but arm-by-arm offset obtained here, and calculate the 1/beta by

1/beta (PA->PB) = (Tpb'-Tpa')*c/L_TOF

, where

Tpb' = Tpb - PBOFFSET(idrun).

PBOFFSET(idrun) is defined as the Gaussian center of the distribution of delta T' (PA->PB).

With this 1/beta definition, go back to T0->PA TOF analysis again.