Gen Software/Simulation Meeting.

Thursday, July 22; 1:30pm - 2:30pm

attendees: Nilanga Liyanage, Andrew Puckett, Bodo Reitz 

1) status of MWDC software

- analyzer able to decode CODA files, and to produce a tree 
  containing the raw-data is running and installed at the UVA PC
- the same software is in the CVS directory at JLab (repository agen)
- some macros/examples to replay the data and to plot some 
  histograms are also installed at UVA
- Andrews tracking code is now compiling and running without
  crashing the analyzer on Andrews and Bodos PC.
  However it prints out lots of warnings, not being able to 
  produce tracks, due to singular matrices. Andrew points outs,
  that some essential geometrical data is missing in the database.
  Also the data-sample is not really good.

2) database for the MWDC

- we went over the database (in the format from July 22) and discussed 
  various entries and improvements.

a) detector map: likes the wire numbers being aligned with the channel
   numbers. However Nilanga points out, that this is not always possible
   (~48 channels per plane are in different order)
   Bodo will look, if he can rewrite the readdatabase code.
b) definitions of the geometry:
   i)   we decided that the origin of the coordinate system will be 
        in the u-plane, at wire #70.
   ii)  the x-axis is in the dispersive direction, particles with 
        larger momentum are at larger x (or in other words positive 
        x is down, negative x is upwards)
   iii) z is perpendicular to the planes, and points in direction of 
        the particles
   iv)  all planes are parallel
   v)   y is perpendicular to x and z, and x-y-z are forming a right 
        handed system.
   vi)  the wire-angle is given in deg, and measured from the positive 
        x-axis toward the positive y-axis.
   vii) the wire-spacing in each plane is given perpendicular to 
        the wires, as is the position of the first wire.
        The wire-spacing can be negative
c) Nilanga will provide the missing parameters
   so far we know:
   z(u) = 0mm , z(up) = 6mm , z(x) = 12mm , z(xp) = 24 , 
   z(v) = 30mm , z(vp) = 36mm

   wire.begin(u) = +-700 mm     (wire numbers start with 0)     
   wire.spacing(u) = -+ 10 mm 
   wire.angle(u) = +- 60deg

d) following items are missing in the database:
   i)  type (u,x or v) needs to be defined, based on wire.angle   
   ii) active area (new)
   iii) angular acceptance (new)
   iv) estimated resolution within one plane
       (new, one could use TDC resolution times drift velocity,
        but it was decided to use an indendent number)

e) following items need improvments
   i) TDC resolution is not used, instead it is defined by the TDC type
   ii) timing cuts need to be checked, if they are used or not
   iii) time offsets: format of the table should be changed
   iv) TTD is not appropriate for our needs, will be removed for now,
       and later replaced by a table in a format chosen by Andrew

Bodo will work on d) and e), and include the parameters c) provided 
by Nilanga. Andrew needs this input, before he can continue.

3) scripts:

the scripts to look at the data need some improvments. Bodo will
look into it. 
A script to extract the T0 offsets is needed. (Bodo will ask Rob)

4) using simulated data:

Andrew pointed out, that the VDCsim class provided from Oles 
summerstudent might be very helpful. It is an example of how to 
read simulated data in the analyzer and treat it like real data.
He will look into it.

5) currently running tests at UVA

We looked at the run from last night. We only found ~150 events
which contained a hit in the MWDC out of 40000. It turned out,
that starting with event 950 the chamber started to say noise.
Nevertheless even before that is only 6% of the events.
Nilange is looking into changing the arrangement of the 
trigger detectors. Right now they are both beneath, it would 
be desirable to have one on top and one below the MWDC.
Nilanga and Vladimir are working on connecting all planes,
removing the noise, checking thresholds etc. However he 
will try to get one long good run over the weekend, so that
Andrew has data to work with.