TRACS

Note

Add the avalanche feature in the TRACS found at https://github.com/CERN-SSD/TRACS .
The purpose of this development is to simulate so-called LGAD, which has "gain layer" beneath the charge collection electrodes. One important policy with this work (so far) is that minimizing the change as mush as possible though re-arrangement of the code structure could fit more in future development.

1. Install

login to cepcvtx.

user@cepcvtx:~$ cd {some directory under which you want to setup the code = Somewhere}
user@cepcvtx:Somewhere$ git clone https://github.com/rkiuchi/TRACS
user@cepcvtx:Somewhere$ cd TRACS

• Necessary libraries, such as FEniCS, ROOT, Eigen, QT4, are assummed to be already installed if we want to run TRACS at the other PC circumstances.

2. Setup

user@cepcvtx:Somewhere/TRACS$ ./setup.sh

3. Build Code

user@cepcvtx:Somewhere/TRACS$ make

4. Run Template

user@cepcvtx:Somewhere/TRACS$ cd ./run/template.IR_Bottom.20190901/
user@cepcvtx:Somewhere/TRACS$ ../../myApp/DoTracsOnly 1 MyConfigTRACS

• Note that the above "DoTracsOnly" executable requests the form as "DoTracsOnly [number of threads] [config file]". However, number of threads == 1, is assumed for this version, ( the original code as well ) and no gurantee if we try to run with more than one thread.

4a. (Optional) Output root files. -- after successfully run "DoTracsOnly" .

You can check several distribution in the ROOT files. Notice that those are bi-products during development/debugging procedures, thus could be changed/modified in future.

• Weighting/Electric fields ("120" is a setting value in the config file and would be different)

user@cepcvtx:Somewhere/TRACS$ root wf120V
root> h_w_u->Draw(); // Weighting potential
root> h_w_f_grad->Draw(); // Weighting field
root> h_d_f_grad->Draw(); // E-field
root> h_d_f_grad_Y->Draw(); // E-field along with Z-axis

• Transient Current ("120" is a setting value in the config file and would be different)

user@cepcvtx:Somewhere/TRACS$ root current120V_scan0
root> i_total->Draw(); // Total current
root> i_init_elec->Draw(); // Contribution from initial electrons
root> i_init_hole->Draw(); // Contribution from initial holes
root> i_gen_elec->Draw(); // Contribution from secondary electrons
root> i_gen_hole->Draw(); // Contribution from secondary holes

• Timing information. Generation time of carriers.

user@cepcvtx:Somewhere/TRACS$ root ncarrier.root
root> e_gentime->Draw(); // Generation time of all of electrons
root> h_gentime->Draw(); // Generation time of all of holes

• Effective Doping Profile along with detector vertical direction.

user@cepcvtx:Somewhere/TRACS$ root Neff_dist.root
root> neff_dist->Draw(); // Effective doping profile

4b. (Optional) "Edge_tree" . -- after successfully run "DoTracsOnly" .

user@cepcvtx:Somewhere/TRACS$ ../../myApp/Edge_tree NOirrad_dt0ps_4pF_tNOtrappingns_dz5um_dy5dV20V_0nns_bottom_0_rc.hetct

user@cepcvtx:Somewhere/TRACS$ root NOirrad_dt0ps_4pF_tNOtrappingns_dz5um_dy5dV20V_0nns_bottom_0_rc.hetct.root

root> .x loadlib.c
root> edge->StartViewer();
or
root> TBrowser a;

• The file name "NOirrad...." would be different, depending on the setting parameters in "MyConfigTRACS"

• The functionality/usage is not fully tested and that's why it is marked as "Optional" now. (09/02/2019)

5. Config File

Here, newly introduced parameters in the config file ("MyConfigTRACS" in the template) are mainly described.

• SetAvalanche = yes # yes: set this doping profile. no: turn-off, even though parameters are given bellow.

The shape of "Gain Layer" is now assumed as a simple Gaussian for which we need to specify some parameters:

• DopingProfile_PeakHeight = 3.0e16 # unit in cm^-3

• DopingProfile_PeakPosition = 1.5 # Z (vertical) position from the top. unit in micron-meter

• DopingProfile_GaussSigma = 0.3 # unit in micron-meter

The (carrier loop)calculation time scales with the number of carriers, including the secondary generated ones due to the Avalanche effect. Thus, stopping the loop calculation iteration, if the number of generated carriers reaches or exceeds the ratio bellow. The ratio is defined as the number between the total number of carriers generated and the number of initial carriers read from a carrier file.

• MaxMultiplicationRatio = 10.0 # no unit.

In addion to above, the effective doping level in the bulk except the "gain layer" is also important. This is determined from the detector thickness and the full depletion voltage, which is already used in the original TRACS. Therefore, you need to set the full depletion voltage correctly. ( In fact, the impact ionization effect happens only at the high doping area, and less effect if one sets the full depletion voltage == nominal doping level wrongly )
One of point is that this full depletion voltage is that one without the "gain layer", to estimate normal doping level correctly.

• DepletionVoltage = 10.0 # in volts

The other factor related with the impact ionization effect is the temperature of the detector. Do not forget about its value.

ChangeLog (from the original repo.)

V1.1)

• Addition of the impact ionization effect (mainly in Carrier.cpp)

• Carrier container is changed so that secondary carriers can be stored as well.

• Note that, above treatment might be dangerous because the memory(stack) overflow can be happen.

• Introducing the effective doping profile option for the avalanche region.

Reference

• Impact ionization effect

The model and parameters are taken from :
M. Valdinoci, D. Ventura, M. C. Vecchi, M. Rudan, G. Baccarani, F. Illien, A. Stricker, L. Zullino, "Impact-ionization in silicon at large operating temperature", SISPAD '99, Sept. 6-8, 1999, Kyoto, Japan.