(from Gene's e-mail)
In principle, I see this as a three step process:
1) determine the shape of SC vs. zdcx: SC = A*(1+B*zdcx)*zdcx fixes B
2) determine what A*GL makes the gaps have no z dependence fixes A*GL
3) determine what value of A makes the DCAs go to zero (and hopefully have no z dependence) fixes A and subsequently GL
That process (where steps 2 and 3 can be done in iterations to refine A and GL and interpolate to a final answer) defines A, B, and GL.
When you quote numbers for SC*GL, I've assumed you've always been quoting the value of GL (the number we store in the DB), and not actually A*GL (a number which is more like 10^-5 or 10^-6 when using zdcx). But you cannot know GL until you know both A and A*GL.
to change values of GL
edit file
StarDb/Calibrations/tpc/tpcGridLeak.20090202.000000.C
change bold line
....
row.InnerGLStrength = 0; // Strength of GL around inner sectors ;
row.MiddlGLStrength = 7.4; // Strength of GL between inner/outer sectors ;
row.OuterGLStrength = 0; // Strength of GL around outer sectors ;
tableSet->AddAt(&row);
....
to change values of SC
edit file
rich/spaceChargeCorR2.20090202.000000.C
change 2 bold lines, use 1st for linear term, 2nd for quadratic term
....
//
memset(&row,0,tableSet->GetRowSize());
row.fullFieldB = 3.6e-7; // Negative Full Field Correction ;
..
row.factor = 1; // Multiplicative Factor ;
..
tableSet->AddAt(&row);
memset(&row,0,tableSet->GetRowSize());
row.fullFieldB = 2.55e-13; // Negative Full Field Correction ;
..
row.factor = 2; // Multiplicative Factor ;
...
tableSet->AddAt(&row);
....