|
Soft Error Research* |
|
The Pennsylvania State University |
|
For Download of SEAT - DA, email rkrishna@cse.psu.edu with your name and contact information. SEAT - DA is a toolset used to estimate the charge collection due to neutron induced soft error. The following fig. 1 outlines the SEAT - DA.
|
|
SEAT—DA |
|
MCNP MCNP is general purpose ,continuous energy , generalized - geometry, time - dependent, coupled Monte Carlo N-Particle (MCNP) transport code used for neutron, photon, electron , or coupled neutron/photon/electron transport, including the capability to calculate eigen values for critical systems. The first part of the scripts involves simulating the n-Si interaction using MCNP. This is done under the script “runMCNP.pl”. This script simulates the structure shown in the Fig.2. The script runMCNP.pl is setup by the setting the structure, energy and the reaction modes. Various reactions are possible like:
|
|
Figure 1. Outline of SEAT—DA |
|
The Scripts used for each of the above mentioned tools are explained below. |
|
The n-Si interactions can have both elastic and inelastic scattering. Currently the inelastic scatterings are modeled. Reactions: The treatment of inelastic scattering depends upon the particular inelastic reaction chosen. Inelastic reactions are defined as (n,y) reactions such as (n, n'), (n, 2n), (n, f), (n, n') in which y includes at least one neutron. Please refer to the MCNP manual for more details. The reactions in MCNP terminology are listed as : "1", "16", "22", "28", "\-2", "102", "103","104","107". This script simulates the n-Si reaction at the input energy levels for various angles and one reaction at a time. |
|
Figure 2: Structure simulated by runMCNP.pl |
|
The following table lists the meaning of all these reactions, see Appendix G:
MT FM Microscopic Cross-Section Description
1 -1 Total
Where MT /FM refer to the MT and FT cards used to generate the results by MCNP. To under stand the MCNP input, the user are referred to the MCNP tutorials available with the MCNP toolset.
Next, “readMCNP.pl” script is used to read the outputs of the MCNP runs. The outputs are stored in a file called “complied results”. This is information should be parsed and given as input to TRIM/SRIM. This output contains the reaction rates for the various reactions and the energies. SRIM/TRIM TRIM is used to find the calculate the charge deposited when a energetic particle hits a crystal. The output gives the charge deposited per unit length. Since, this tool is based on Windows platform. The energy deposited tables for the various ions generated by MCNP are generated and stored on text files. Next “let.pl” script parses these and produces the files in the format the device simulator Davinci understand. These files are called as the LET files.
Last tool in this toolset is Davinci. Davinci is a device simulator. First input at this stage is a device model, next the LET files from the SRIM outputs. Lastly we need to input the geometry of the charge column. Using this inputs we can simulate the single event upset. Two scripts are embedded here in perl get the job done. First script inSEU.pl, setsup the scripts for Davinci and simulates the input files. Next script, pmosCurve.pl generates current waveforms from the outputs.
* This work is supported in part by NSF Award # 0454123 |