Main INDEX
Monthly INDEX
PREV
NEXT
User name R. Michaels
Log entry time 17:35:55 on October13,2000
Entry number 50311
keyword=Origin of TC wierdness, poss. fix
Regarding the TC wierdness in halog 50304,5, and 7.
First, I assume that Jeff's 'tc' has the sign convention
(opposite to what I'm used to) that bigger 'tc' means
the Left spectrometer is earlier. Then, the region from
192 to 208 in his raw 'tc' corresponds to Left arm so
early that the T5 timing is carried by R-arm and the
DAQ ``L1A'' pulse also is carried by R-arm. Then at
the L-arm retiming circuit the L-arm retiming pulse
can be too early (see halog 49249 and paper log book
diagram on about Oct 4). Result is that L-arm gates
become tied to R-arm scintillator. Very ugly. It is
related to our delay spools being too short. We made
this region of bad timing worse when we increased the
randoms window (actually when we increased the width of the
L-arm signal, the T3 trigger, see halog 50013, paper
log book p. 179). So, if all this is correct the options are:
----
1) If the data from raw tc=192-208 (Jeff's sign convention)
contain no important information, you can cut it out. I
think the rest of 'tc' region is ok.
2) We could add some more delay cable to the retiming
circuit on L-arm. A ~16 nsec cable would suffice. However,
Jeff's plots of ADCs and TDCs shows that when we shift 15 nsec
we begin losing ADC amplitude. This is exactly the problem I
have unsuccessfully fought... delay spools.
3) If we reduce the L-arm trigger width to 8 nsec (which I
DO NOT recommend reducing it below ~15 because it may
not fire modules), it would also get rid of this funny region.
-----
My vote is 1), and someone should double check me.