User name R. Michaels
Log entry time 17:03:51 on October 3,2000
Entry number 49204
keyword=Coinc Time Meas -- NOTE: T0 Change !During day shift I looked at random coincidences and
pulser coincidences and noticed problems with the coincidence
time measurements. The coinc time is measured on both
spectrometers using a TDC 1875 with 0.1 nsec resolution.
L-spect: roc2, slot 23, channel 48 (channel number starts at 0)
R-spect: roc1, slot 21, channel 4 (channel number starts at 0)
Traditionally 'TC' is the latter (called 'spare_4' in detmap),
but it may actually be more accurate to use the L-spect now
because I think it uses less signal delay, hence less jitter.
-----------
Problems:
1. L-arm cabling problem (was using R-arm gate for stop, should
use R-arm trigger instead).
2. R-arm fastbus gate timing tied to the event type instead of tied to the local
scintillator. This was because some bit was off in a module. The result
was that for true coinc the fastbus start timing was determined by the
L-arm which also makes the stop -> delta function (hard to spot with
pulser). Now the start comes from the R-arm, the stop from the L-arm
as required.
** IMPORTANT NOTE: T0 on the R-arm changes !!!
Previous database measurements will not work for subsequent data !!!
Below are 'TC' on both spectrometers. As hadron momentum decreases
I expect the peaks to move: 1) R-arm reduces, 2) L-arm increases.
Hints: 'goonlana', run ha_analyzer, open g_hist_run#.hbook, use coinc.kumac,
see 'spare_4' and 'spare_5' in detmap_nofpp.config
FIGURE 1
