Sphenix-tpc-l mailing list

["Huang, Jin" <jhuang AT bnl.gov>]

Dear TPC software team

 

Could you help me to understand why the tpc cell size in the z-direction is 0.17/2 cm?

 

This question comes in two parts:

1.       The z-size (as variable tpc_cell_y) is set from 0.17 cm to 0.17/2 cm when changing from 60->40 layer TPC simulation. May I confirm the consideration here?
https://github.com/sPHENIX-Collaboration/macros/blame/master/macros/g4simulations/G4_Svtx_maps_ladders%2Bintt_ladders%2Btpc_KalmanPatRec.C

2.       The original number 0.17 cm was described as “z resolution * sqrt(12)” as in
https://github.com/sPHENIX-Collaboration/macros/blame/4275178171858a7a491d19e44166a2285c6602e5/macros/g4simulations/G4_Svtx_maps_ladders+intt_ladders+tpc_KalmanPatRec.C
This implies the originally expected z-resolution should be 500 um. This is clearly not the case as in Tony’s results, which has a z-resolution of 180 um (that good?).
>From the electronics point of view, the z-length/ADC sample = 1/(10 MHz) * (6 cm/us) = 0.6 cm. Since we are not yet simulating the FEE shaping timing, which dominate the z-size of the cluster (rise time ~160 ns), an effective z-sample size could be different from 0.6 cm. But I would be surprised that it would be one order of magnitude smaller (0.17/2 cm).

 

As the TPC simulation evolves to a more realistic setup (single active volume as presented by Carlos), I guess we just need to set the legacy TPC simulation with cylindrical-layered approximation to roughly reproduce the resolution and occupancy, and make it work for the HIJING simulation users. It will be great if you could help me understand whether the current z-binning would be too small. A reference to previous slides would be useful.

 

As the z-cell size directly affect the memory usage and the z-resolution of the TPC track, it is quite important to get it right.

 

I would also suggest to introduce the shaping time into the current simulation (double sigmaL = 0.48; // 80ns shaping time in sigma * 6 cm/us as in PHG4CylinderCellTPCReco::process_event()), and set z-cell size to 0.6 cm (1/(10 MHz) * (6 cm/us)). Then it further reduces the cell memory consumption to 1/3 and would produce proper z-resolution and occupancy.

 

Thank you,

 

Jin

 

 

______________________________

 

Jin HUANG

 

Associate Physicist

Brookhaven National Laboratory

Physics Department, Bldg 510 C

Upton, NY 11973-5000

 

Office: 631-344-5898

Cell:   757-604-9946

______________________________

 

From: sPHENIX-tracking-l [mailto:sphenix-tracking-l-bounces AT lists.bnl.gov] On Behalf Of Anthony Frawley
Sent: Wednesday, July 12, 2017 12:26 AM
To: sphenix-tracking-l AT lists.bnl.gov
Subject: [Sphenix-tracking-l] Followup on increased TPC cell size

 

Hi All,

 

Following Haiwangs's presentation at the simulations meeting today, I made some performance plots for tracking with the TPC cell size doubled. I posted them on the simulations meeting page:

 

https://indico.bnl.gov/materialDisplay.py?contribId=2&materialId=slides&confId=2701

 

The file name is:

 

double_tpc_cell_size_tests_july11_2017.pdf

 

The bottom line: the performance is very good. The TPC simulation parameters should be re-optimized carefully to match the larger cell size, but this reduces the standalone tracking memory usage for a central Hijing event from 15-20+ GB to less than 7 GB.

 

Cheers

Tony