Proposed HV correction for FGT aiming to equalize response, reduce overall gain to move away from the saturation peak. 

Based on some of HV scan runs taken on February 9, the pulse time-shaps have been  identified with the 2 threshold algo: 

pulse-shape filter:  minAdc<100, maxADC>200, minBin=0,1,or 2

Fig 1. show example of time-pulses, the maximal value of the pulse was used in this analysis.

All strips belonging to given quadrant were processed. DB was used to subtract ends and get rid of bad strips. Additional private mask removed certain APVs.

Slopes of the spectra were fitted with exponential in the range  above pedestal and below saturation peak (ADC range [500,1200]).

Relative gains is defined as -1/slop.

Fig 2. Shows  comparison of inclusive spectra for the quadrant 1B, for 5 HV values. Note, the stats vary sets to set. The relation of gain to HV is shown in the bottom right, straigh line fit is drawn. All 14 quadrants are in teh attachment.

Fig 3. Summarizes gains for all 14 quadrants determined for 3400V and lists proposed new HV which should lead to the peak of ime shape to not exceed ADC of 1000.

Fig 4. comparison of spectra change after HV adjustment, looks better

Fig 5. Spectra for new HV set 2, 43k input minB events . HV

Fig 6. Spectra for HV set 3, 65k input minB event.

Fig 7. Gain's form slopes HV set 3, zoom in form above.