Residuals of global tracks vs. padrow for each sector in Run 12 pp510, for 6000 events from run 13108073 (see RunLog for Online QA), using Jan & my calibration of GridLeak dependence on anode voltage, and several cuts on the global tracks used to improve quality and reduce pile-up:

• has primary track partner, whose primary vertex is...
  • the highest ranked in the event
  • has a rank greater than 0
  • has at least 3 BEMC matching daughters
  • has at least 3 BTOF matching daughters
  • |v_Z| < 50 cm
• doesn't cross the TPC central membrane
• |η| < 0.8
• 0.5 < p_T < 10 GeV/c
• N_hits ≥ 25
• and hits from the track used for the residuals are...
  • within |φ_local|<0.2 of sector center (sectors span |φ_local|<0.262)
  • within |z|<200 cm (i.e. are not "prompt")
  • are on the correct size of the TPC for the sector which recorded them (otherwise would be known to be pile-up)

Dead channels obervable in the Online QA:

• RDOs: 5-1, 6-1, 7-1, 14-3, 21-1, 22-2
• Anodes: 3-4, 20-5, 23-1; 7-6 is also reduced, and there are the usual oddities on 5-8 at padrows 38-40

Note: you can open each image in a new browser tab or window for larger PNG, or click to view PDF.

The first plot is average over all z (vertical axes are [-0.075,0.075]):

Next is the a plot of (Resid₁₄ - Resid₁₃)/2 as a function of z (this is essentially half of the residual gap from padrow 13 to 14), averaged over all sectors excluding sector 14 and sectors 2,3,4,20,21,22 (i.e. |φ_global|<0.786):

______

Below are a pair of plots where I first tried to make these plots but accidentally required CM crossing instead of excluding it, which means that all these tracks are at small z and convolute east and west distortions distortions a little. I did not re-run the exclusion of the first two padrows for the above plots.

Residuals after excluding the first two padrows from tracking: 

-Gene