Make New Entry, Make Followup Entry
User name R. Michaels
Log entry time 06:25:25 on November12,2010
Entry number 339671
Followups:
keyword=helicity problems gonePlotted in fig 1 is the time lost and fraction of data lost due
to "confused helicity". Note, I've fine-tuned the gate-off time
to make the sum-rule for ungated minus sum-of-helicity peak at zero.
This was also done for the end-run summaries during this owl shift.
For the 209 runs of duration 5 min or more since Friday Nov 5 at 17:00 the error
rate was better than previously but not perfect, as was noted in the halog
a few times since then. On average 11 seconds (0.3% per run) lost.
These are the top 2 panels of fig 1.
Yesterday, Julie and I made some more improvements. In the 30 runs since
then (14 long runs) we've lost far less than 1sec (fraction < 1e-4) per run, see
the bottom 2 panels. Note the very narrow range of X-axis and
the delta-function peak on this range.
Note, not every run records to scaler_history.dat. Run must not be a CODA crash.
The preliminary conclusion is that the helicity errors are gone.
FIGURE 1
