QASY: Charge Asymmetry Measurement & Feedback
Robert Michaels, rom@jlab.org, Jefferson
Lab Hall A, updated July 2006
(NOTE added Nov 2008: Obviously we don't use the HRS scalers for feedback during Transversity, although we did that in the past. Transversity uses the HAPPEX DAQ for feedback. Still, one may try to run qasy to check the qasy.)
To measure the charge asymmetry from Hall A HRS DAQ scalers
use the code "qasy''.
This code generates an ntuple file "scaler.root"
from which a number of other analyses of scaler
data are possible.
Some root macros exist to facilitate this.
If the beam is quiet, the
charge asymmetry has an accuracy of 10 ppm for 1-hour CODA run.
RUNNING QASY
- Login to adaql4 as "adaq" account. Go to the
correct directory which is ~adaq/qasy
- Type "qasy run-number" (without the quotes).
E.g. "qasy 1354" for run 1354.
- If you leave out the run number it will analyze
the most recent run.
- More option may exist: Type "qasy -h" for help.
- After 0.1 to 5 minutes (depending on filesize), a ROOT plot pops up with results.
- Results are the charge asymmetries on two monitors (upstream `U' and downstream `D') measured with two gain factors (3 and 10) using the V-to-F's and scalers.
- At the root prompt, you are invited to type ".x asyfit.macro". (dot x)
- asyfit.macro obtains a fit to the asymmetries which is also printed on the screen where you have the root prompt.
ADVICE about Qasy script
- There are beam cuts. I think they are about 0.5 uA now. You may type ".x beam.macro" after running
qasy, and see if the cuts in "asy.macro" are correct. Also if the
beam is very unstable, the fits may not work too well because of tails, etc.
- It might be a good idea to at least once induce a deliberately
large charge asymmetry and compare it for example to Moller. A host
of evils may produce bad results, including scrambled helicity information
and mistimed gates.
- There are a few more details in ~adaq/qasy/README. This software
does more, of course, than just analyze the charge asymmetry. An ntuple
is created with many variables from scalers.
This page maintained by
Robert Michaels.
rom@jlab.org