From jmichaelfinn@cox.net Thu Sep  9 07:23:14 2004
Date: Tue, 07 Sep 2004 12:02:01 -0400
From: John M. Finn 
Reply-To: finn@physics.wm.edu
To: 'Dave Mack' , 'Roger D. Carlini' 
Cc: 'Allena Opper' , neven@phys.latech.edu, pitt@vt.edu,
     tony@phys.latech.edu, grimm@jlab.org, finn@jlab.org, jmammei@vt.edu,
     nmorgan@vt.edu, smithg@jlab.org, birchall@physics.umanitoba.ca,
     armd@jlab.org, yl0094a@jlab.org
Subject: RE: Qweak working group - Comments

Dear Dave and cohorts,


I would be willing to augment my code to look at these sensitivities. By
using the same random seed for small offsets in angle, etc, I can extract
the first order acceptance averaged sensitivities in a straight forward
manner. 

However, I am missing two critical elements to complete this analysis:

1) I need a ray-trace mapping of the spectrometer transport. Hopefully Klaus
can get that for me, if someone else hasn't already generated one.

2) I need the light weighted output of the Cerenkov verses incident position
and angle, including edge effects. (Can anyone get this for me?) 

I agree with Roger: Now that we have established that an appropriate
collimator geometry restores the baseline of the experiment, we should use
caution in making major changes without vetting their potential collateral
effects.

By the way, I found and corrected an error in my estimate of the uncertainty
of the B term. The units didn't match (oops) but fortunately the numerical
correction was small, so the conclusions I presented remain valid. 

Regards to all,
Mike

-----Original Message-----
From: Dave Mack [mailto:mack@jlab.org] 
Sent: Tuesday, September 07, 2004 10:58 AM
To: Roger D. Carlini
Cc: Allena Opper; neven@phys.latech.edu; pitt@vt.edu; tony@phys.latech.edu;
grimm@jlab.org; finn@jlab.org; jmammei@vt.edu; nmorgan@vt.edu;
smithg@jlab.org; birchall@physics.umanitoba.ca; armd@jlab.org;
yl0094a@jlab.org
Subject: Re: Qweak working group - Comments

	Roger,


On Tue, 7 Sep 2004, Roger D. Carlini wrote:

> 
> Hi Folks:
> 
> My understanding of where we stand is the following:
> 
> 1)  We have re-established a base design which will allow us to achieve a
4% qweak measurement in about 2200 hours + 
> comissioning time (at 85% polarization) even after a deduct for internal
radiative effects. There are infact several 
> solutions to choose from at various are Q**2 and Theta acceptances which
fit on a slightly longer than 2 meter bar.
> 

	Name the baseline and I'll try to estimate, using reasonable
alignment tolerances (target relative to collimator), the errors on ,
, and the net impact on Qweak. At the time the proposal was written,
it was assumed the collimator would be far enough away from the target
that longitudinal alignment tolerances would be trivially achieved.



	Dave