The TPC GridLeak distortion is understood to be due ion generated in the high gain region near the TPC anode wires leaking out into the main TPC volume via the gap in ground (cathode) and gated grid wires between the inner and outer sectors. It is believed that both the inner and outer sectors contribute to this, based on observations first made in 2006, which further demonstrated that the anode voltage used is also a factor in the amount of distortion.

A calibration of this has recently matured.

To allow for future updates up this calibration, and handle reproducibility of past productions, it is appropriate to put the results of this calibration in the database.

I propose a table with 24 parameters, representing the 24 degrees of freedom of each inner and outer TPC sector contributing differently to the GridLeak distortion, and described in the following model:

  Distortion metric = SpaceCharge * GL * Amplitude[sector] * (IOratio[sector] * RelativeGain_Inner[sector] + (1-IOratio[sector]) * RelativeGain_Outer[sector]) 

...where SpaceCharge is already designed to vary with luminosity. The dependence of SpaceCharge and GridLeak on luminosity is already calibrated with each collision species, energy, magnetic field, year of operation, and database tables already exist for these. The past model has been to use only:

  Distortion metric = SpaceCharge * GL

...and fold in any changes in gain into the calibrated GL (as GL was recalibrated often anyhow). The problem with this approach is that any time the gains are changed (via anode voltage changes), the calibration must be re-done. Further, it does not compensate for altered voltages on only a subset of the TPC sectors, which has been an issue since day 1 (sector 20), but has become an increasing issue in recent years where anode voltages have been reduced to prevent tripping of the TPC power supplies.

So, my proposal is a new table under Calibrations/tpc/tpcGridLeakSectors, which would have 12 elementIDs for each of the 12 TPC sectors. This table would rarely need updated, perhaps having one entry (per elementID) per year of operation in general (though compatibility with past years may require a few multiple-entry years). 

/*    tpcGridLeakSectors.idl
*
*  Table: tpcGridLeakSectors
*
*       description: compensate GridLeak distortion for individual sectors
*/


struct tpcGridLeakSectors {

   double         Amplitude;       /* amplitude of sector's contribution to GridLeak at design V */
   double         IOratio;         /* ratio of inner sector contribution to outer at design V */
   char           comment[255];    /* considerations in determining values */

 };

-Gene