User name R. Michaels
Log entry time 07:03:15 on November 8,1998
Entry number 4891
This entry is a followup to: 4860
keyword=Study of systematics (HAPPEX DAQ)Here I summarize the overnight runs with HAPPEX DAQ
and comparison to e94010 data.
The helicity correlated beam position differences were small,
less than 1 micron, typically only 1-2 sigma. Log entry 4881 is
a typical run. Assuming a reasonable power law for the energy
and angle dependence of the cross section, the false asymmetries
should be less than 100 ppm.
The charge asymmetry drifts around slowly and is presently about
-300 ppm. Since the "PITA slope" is fairly stable and the drifts slow,
a feedback loop could control this.
Comparison to "pscaler" asymmetry is in the plot below. The HAPPEX
data points are stars, and the e94010 data are open circles (errors unknown).
Sometimes there are large discrepancies, and I think I know why. For the
runs indicated by an arrow, the beam tripped off a few times. For the other
runs the beam was remarkably stable. If the beam trips off, the naive "i10"
asymmetry develops errors of order 1/T, for 1-second helicity period, where
T= time of run (seconds), so the error is of order 300 ppm. You'd have to be
lucky for the beam to be off equally for each helicity. This can probably be
reduced by looking at the 2-second updates of scaler data in the data file
(Wang will do this). For the runs with stable beam there is a systematic shift between e94010 and HAPPEX. Not sure why, but it may have partly to do
with the scaler offset for zero current, which is not used in pscaler. It would
be interesting to see this shift removed.
FIGURE 1
