From carlini@jlab.org Thu Sep  9 07:28:57 2004
Date: Tue, 07 Sep 2004 16:36:14 -0400
From: Roger D. Carlini 
To: Neven Simicevic ,
     Jim Birchall , Dave Mack ,
     Juliette Mammei , Klaus Grimm ,
     armd@jlab.org, Mark Pitt , Allena Opper ,
     Greg Smith , Norman Morgan ,
     Mike Finn 
Cc: Yongguang Liang , Shelley Page ,
     Tony Forest 
Subject: Re: Qweak working group - next meeting

Hi Folks:

If you recall the figure I sent around before the last meeting showing long (end overlapping) bars that allow for full 
phi coverage. Such an arangement would work nicely with Jim's options 12-8-22, 12-8-23 and 12-8-24. There would still be 
enough space for supports between the inner and outer concrete shielding.

Another possible modification would be to use 2 bars/octant (which would probably be required anyway due to the total 
lengths), but rather than glue them together - mirror and overlap the ends in the center of each octant. This would 
result in a 16 fold symmetric detector. The 2 signals could be added together electronically if we wish. A small 
correction to the statistical error would need to be made to account for the overlap.

The bars could then be tilted slightly reducing their required width. This arrangement might allow the use of "narrow 
bars" readout from the end and no glue joints. The hours would now be ~1500 (per Jim's table) for a 4% measurement 
assuming the "B" term is under control.

The "slick" feature of this configuration is that we can pretty much select any "phi1-phi2-phiphi" (ie. average Q**2) 
selection from Jim's table by constructing addition Cu-Pb (4" to 6") single collimators at ~1.4m from TGT. As in Jim's 
model the front disk would become (like the clean up collimator) just a water jet cut non-acceptance defining "shield" 
for the GEMS.

This all ignores issues about the chamber dimensions, but our existing chamber designs should work for Jim's nominal 
"phi1-phi2-phiphi" configurations. We migh be able to get coverage for the larger acceptances by taking two measurements 
per octant instead on one.

Best Regards

Roger