W&M  report 08-12-2004


1) Clean room status

We managed to take down the heavy duty light housings without risking a collapsing ceiling tile support.
After that the ceiling was painted with a concrete ceiling paint in order to avoid dust and flakes released from
the ceiling.

2) Garfield simulation status

Our summer student Carissa Capuano is working on the drift time to drift distance decorrelation after she
managed to fit the XT-data (simulation results from Garfield). In the beginning the fit resolution was poor
because of a coarse spacial track stepping size in the simulation. After we repeated the simulation with tracks
with an given angle range and a displacement stepping size of 200um we could resolve the beginning of the
avalance region with much better resolution. The drift time->drift distance decorrelation depends strongly
on the definition of the fit regions (close to wire: quadratic, else: linear).
Up to now Carissa managed to resolve a simulated track within a precision of 100um which is much better than
the track resolution of ~250um obtained in a  MIT thesis for the Hall A VDC using the same decorrelation technique.
  
(Plots will follow ... Maple's Postscript files are unreadable by other programs ...)

3) VDC design status

In the previous report I observed that the lower corners of the front chamber are exposed to the ajacent octants
(see pdf here). Therefore I had to reduce the total width of the front chamber by ~20cm. The pic below
shows a top view of the VDCs. The semi-transparent wedges are the electron envelopes including a clearance
zone of 1cm in each direction.




Recently we discussed the number of drift cells beeing hit for various track and VDC constellations.
Whatever (regular) track crosses the VDC, the responding number of drift cells should not be lower
than 3 and higher than 8 . The more drift cells are hit, the better the track reconstruction.



The angle shown in the upper picture is NOT the incident track angle. The incident track
has to be transformed into the perpendicular wire plane in order to extract the number of responding cells.
Using a local coordinate system defined by the VDC frame, the track angle seen by the perpendicular
wire plane is quite complicated:


This is the actual angle which will be extracted by any drift chamber tracking program in a experiment, e.g.
in Garfield this is the tracking angle to be used for simulation.
Theta and Phi are the in/declination track angle, Gamma is the angle of the wires in respect to the frame.
A more detailed drawing, angle definition and derivation can be found here (nice write up will follow).
As a start up I have written a small Maple program (tracking_angle.mw) that  calculates the  number of drift
cells beeing hit depending on the input parameters: track angles Phi and Theta, Wire string angle, wire distance and
the distance of HV to wire plane. The results for selected input parameters can be found here (Track_Calc_Results.pdf)

Summary:

For optimizing the tracking efficiency the ~polar angle is quite large at the ends of the drift chamber
due to the defocussing of the QTOR magnet (see right pic below).

With a 45deg wire string angle a ~polar angle of 0deg  the U and V plane will respond with with the same
number of drift cells beeing hit. For a ~polar angle of 20deg this number will increase for e.g. the U-Plane
(5.91 hits@45deg) but will decrease for the V-Plane (2.76 hits@45deg).

After some tweaking we get a better and more balanced (min,max) response when we use
the following settings (see the last two section in Track_Calc_Results.pdf):

U-Plane wire string angle: +30deg
V-Plane wire string angle:  -30deg

Distance HV plane - Wire plane : 13mm    (half drift cell height)
Distance wire - wire                    : 4mm      (corresponse to 8mm spacing on the frame)
                                                 (was 4.23@45deg wire angle for Hall A)
                                                 (corresponse to 6mm spacing on frame)  

This will lead to 300 (Front Chamber) or 325 (Back Chamber)  signal wires for a single plane.

In total we have for 2 dual VDC:  2500 signal wires.
 
Due to the delay line readout schemata we need only #planes*#interleave*2= 8*8*2 = 128 TDC channels

Channel cost: ~$65  (TDC + Preamp),
-> Total channel cost: 128*$65 = ~$8,500 for 2 Dual VDCs

 The next step will be to include a dummy VDC into Geant and extract the number
 of responding drift cells and compare it to my first order calculation.


        sideview: azimuthal angle                              topview: polar angle


4) Roger's alternative drift chamber proposal

 ... will certainly be discussed today

Design data for the rotating HDC proposal

Inner size: 490mm x 180mm, active area: ~490mm x 160mm

Number of wires per plane, assuming Region2 design:

U,U',V,V' :  134 wires each, 67 signal wires each
X,X'         :   98 wires each, 49 signal wires each

1 Single HDC   :       732 wires total,    366 signal wires
2 Dual   HDCs  :     2928 wires total , 1464 signal wires

Total Channel cost:  1464*$65 = ~$100,000 for 2 Dual HDCs