From mack@jlab.org Thu Sep 9 07:27:01 2004 Date: Tue, 07 Sep 2004 15:49:45 -0400 (EDT) From: Dave MackTo: Jim Birchall Cc: Roger Carlini , Mark Pitt , rams@triumf.ca Subject: Re: Qweak working group - next meeting Jim, If this is a valid point (I'm probably on your side), it should be consistently applied to the entire experiment (eg, what is our specification for beam motion at the line frequency with ANY collimator scenario? assuming that ripple is dominated by the line frequency). I believe the CEBAF tradition is to assume that effects at the line frequency are perfectly rejected. But since we slip phase between the reversal and the line frequencies, the story can't be that simple. Roger assures me the end result will be the same as long as the experiment is perfectly linear. Des said something similar at the last collaboration meeting. Clearly, this can only be true if the ripple amplitude is stable over "enough" power line cycles (ie, that the ripple intensity noise be negligible). A rigorous mathematical treatment would help define "negligible" as well as other hidden assumptions. regards, Dave On Tue, 7 Sep 2004, Jim Birchall wrote: > Dave, > > How about noise on the asymmetry measurement? How much ripple is > allowable on the mini-torus field so that the noise is less than say > 10% of counting statistics on Qw? A small change of deflection of > elastic electrons could modulate the event rate nicely, especially at > the small angle side of the acceptance where the cross-section slope is > large. > > Jim >