User name R. Michaels
Log entry time 23:06:18 on January11,2005
Entry number 139024
keyword=ctimer problems (L1A too late)The plot of ctimel (Y-axis) vs ctimer (X-axis) looks like fig 1.
The lego version is fig 2.
There are two problems. One is that the L-arm coinc time (ctimel)
reaches to ~3500 channels which is near the end of the range and
causes a peak. But this is ok since the trues are at chan 2500.
The 2nd problem is that the R-arm coinc time (ctimer) has a delta
function, starting just above the trues, which occurs because the
L1A is too late. It would be easy to take access and remove 20 nsec
from my copy of L1A that goes to R-arm, if I have 20 nsec to remove
(not sure). However, we decided to run like this, it just means we
lose some redundancy. The R-arm is a little harder since the T5
timing is tied to T3 and this slides around wrt the retiming and
the window is only 200 nsec.
FIGURE 1
FIGURE 2
