Main INDEX
Monthly INDEX
PREV
NEXT
User name R. Michaels
Log entry time 00:37:47 on January 7,2003
Entry number 90653
This entry is a followup to: 90646
keyword=ESPACE helicity and DAQ
I'm not in a position to check the ESPACE algorithm myself, but I have
written the C++ equivalent of this for the new analyzer and used that just
now to investigate this problem. Note, it is known that the ESPACE algorithm
is superior (I haven't worked as hard as Xiadong) but the overlap is >= 98 %
so it's a rough check.
The problem with the Right arm data is the nature of the deadtime -- apparently
the DAQ "dies" for anomolously long periods of time, or order 1 second. The
probability of this is sufficiently high that my C++ code throws out most data.
In the meantime, I analyzed the scaler data (which has zero deadtime) and see
that the G0 helicity structure is PERFECT for both Left and Right spectrometers.
So, there is hope that when we cure the Right arm DAQ disease, ESPACE will
work fine. Meanwhile it should be checked on Left arm (someone, please ?).
Note, there should be no problem with helicity due to deadtime -- in principle --
because whether the deadtime is high or low, you should see all transitions since
the probabilty to be dead for all of 30 milliseconds (one helicity pulse) is
negligible *IF* the deadtime is normal. Typical DAQ deadtime is 200 microseconds
and exists <= 35% of the time. But if the DAQ skips several whopping seconds,
you may have trouble reconstructing the helicity reliably. One help is that you can
corroborate the 100 kHz timestamp using the 2nd timestamp (1024 Hz) available
in scalers.