Negative Transconductance - Effect of Lattice Heating : Negative Transconductance - Effect of Lattice Heating

Requires: S-Pisces/Giga
Minimum Versions: Atlas 5.22.1.R

This example compares the results of Ids/Vds curves for an SOI transistor using isothermal and non-isothermal approaches. The heating of the silicon film causes a negative saturation slope. The file shows:

  • SOI structure formation using Atlas syntax
  • Selection of the lattice heating model
  • Selection of thermal boundary conditions
  • Id/Vds simulation for Vgs=10V
  • Repetition of the Ids/Vds curve excluding lattice heating for comparison

There are two Atlas runs in this example file. The first uses lattice heating and the second uses isothermal models.

The initial phase of the input file constructs an SOI device using the Atlas syntax. mesh, region and electrode statements are used to define the geometry and mesh of the structure. The doping statement is used to define analytical doping profiles. A uniformly p-doped silicon film is used with heavy gaussian n+ source and drain regions.

The heat flow equation is selected using models lat.temp . Other models such as arora are used to define a temperature dependent mobility model. Local lattice temperature is used in the documented mobility model equations. The thermal boundary conditions are crucial in simulations with lattice heating. The thermcontact statement defines the heat sink. The heat sink is on the bottom of the silicon substrate region. All other boundaries are presumed to be in thermal isolation.

The choice of numerical method in SOI device simulation is also important. The standard newton method has some convergence problems in floating body devices. As impact ionization occurs the initial guess to the potential of the floating channel region becomes more difficult. The newton method, which relies on a good initial guess, can encounter convergence difficulties. These difficulties are generally not fatal to the simulation, but require small voltage steps to be used. This adds to the CPU time considerably. A more robust method available in Atlas is by choosing the combined gummel/newton algorithm.

In addition, the solution of the heat flow equation must also be included. The most robust method, once the temperature rises above the heat sink level, is newton. However in the early stages of the solution the decoupled block method is better. Atlas can switch between these methods so all are included in the method gummel block newton statement.

The electrical solutions for Ids/Vds is defined in a similar manner to the regular NMOS example described in the MOS examples section.

The simulation with lattice heat is exactly analogous to the subsequent one using isothermal models. The run differs only in having no lat.temp or thermcontact definition.

Id/Vds curves form the two runs can be overlaid in TonyPlot showing the effect of lattice heating on the device performance. The increased temperature causes the mobility to decrease as the drain bias rises leading to a negative saturation slope. Comparison of the physical variables at the same bias is also possible from the saved solution files. Plotting impact ionization in each case shows how the elevated lattice temperature reduces the impact ionization rate.

To load and run this example, select the Load example button in DeckBuild. This will copy the input file and any support files to your current working directory. Select the run button to execute the example.