Main INDEX
Monthly INDEX
PREV
NEXT
User name R. Michaels
Log entry time 11:07:23 on April 23, 2007
Entry number 199741
keyword=timing checks for T5 and T6
There was a concern about D.ctimel for T5 and T6. If I understood
correctly, the idea was that for high Qsq there is no split between
T5 and T6 in this variable, but at lower Qsq a split of 20 nsec was
observed. Since this might grow even more and (more importantly) since
it should not grow at all, it indicates a marginal timing problem.
A possibly seperate problem is the observed L.tr.y vs L.tr.x. I didn't
do the analysis myself, but what I've been shown suggests T5 disappears
on one side of the focal plane while T6 continues to exist, and then
T6 disappears on the other side while T5 continues to exist. This
should be confirmed or denied by careful analysis because it is
extremely worrisome if true. Actually, I really doubt it.
I tried to reproduce this with our pulser simulation. Plugged in
the usual 50 nsec cable for L-arm S2m, and did runs with only T5
and other runs with only T6. I observed that the retiming picture
was extremely stable: L1A arrives 45 nsec before the RT pulse on
L-arm for each trigger and for the full range of beta. This is
how it should be. Note, for low Qsq there are 20 nsec diffs between
one side of the focal plane and the other on R-arm. This is a little
tight compared to 45 nsec, but to make it less tight would require
more analog delay (remember all the trouble to reach beta=0.21).
Next I looked at the D.ctimel on the scope for several beta values.
It was very stable (124 nsec) except at lower beta where it walked
to 100 nsec because the EDTM modules are not able to keep up (I have
a problem that the EDTM doesn't have enough range, something I'll fix
this summer). But D.ctimel was always in the range of the TDC for
the pulser.
The conclusion is that I don't see a problem with the trigger except
that L-arm RT is tight. But we already knew that.
A copy of this log entry has been emailed to: petratos@jlab.org,camsonne@jlab.org