When an implant occurs there will be some level of damage to the crystal structure of the silicon. If this damage is not annealed out then any subsequent implant will have a different penetration depth compared with the crystal that has no damage. The reason this occurs is due to the crystallographic nature of silicon. This means that there are some “channels” along certain crystallographic directions where ions can move much more freely.
Implementation of a new ferroelectric capacitance model from Ramtron International Corporation into SmartSpice was first described in the April 2002 issue of Silvaco Simulation Standard. This model utilizes a new concept of double distribution of domain reversal voltages. The temperature effects were not detailed in the previous article. This application note discusses the implementation of the temperature effects and updates the device syntax.
Version 7.0 of the University of Florida Silicon-On-Insulator (UFSOI), released in 2002, is now available with Silvaco SmartSpice by setting LEVEL to 21. SmartSpice uses version 7 by default, but versions 4.5, 5.0, 5.0 rev 1.0, and 5.0 rev 6.0 are still available through resetting the VERSION and REVISION parameters.
Silvaco SmartSpice simulation results are typically stored in RAM. Since transient simulations of large circuits often exceed 1GB, a large swap-space partition is required prior to simulation. Constant disk access may dramatically decrease simulation speed, therefore shifting some or all of the load to the system’s memory helps to alleviate this problem.
One of the methods to define a Voltage Controlled Oscillator (VCO) in a SPICE simulation is to make use of a voltage-controlled voltage source (E element). A typical VCO expression makes use of a sinusoidal function as shown in Formula 1.
Level=36 RPI poly-Si TFT and other original TFT models do not support automatic geometry binning. The enhanced model included with SmartSpice 1.9.7.C or later now supports the selecting of a suitable model with both LMIN/LMAX and WMIN/WMAX, as well as with other MOSFET models.
Silvaco Expert is an extremely powerful layout-editing tool that supports features related to parameterized cells. Parameterized cells, often called P-cells, increase designer productivity by adding enormous flexibility and efficiency to the design process. While standard cells help the designer to avoid repetitive drawing of identical pieces of layout, P-cells extend this functionality to the specific parameters that define the mask geometry. As a result, P-cells assist in the automation of layout design and help speed-up modification through the revision of P-cell parameters instead of wasting valuable resources by repeatedly redrawing the layout geometry.
Modern complex circuit designs require automatic router functionality in circuit layout editors. In multi-layer hierarchical design, it is often necessary to build a single wire between two objects or arbitrary points. It is unpractical to call a heavyweight router that is hardly aware of the design settings and technology parameters. Expert’s built-in Wiring Tool helps to bypass these time consuming operations, since all necessary information for routing is made available during Expert session.
The C++ interface for Silvaco’s Expert layout tool, is now an exclusive customization language that expands the tool’s handling of high-level customization, such as a design application working with Expert’s database. Users of the interface develop dynamic link libraries (DLLs) with Microsoft Visual C++ that are stored in the Silvaco install directory. Expert loads these DLL files and implements the user-defined commands in the menu bar.
After I create my round curve and run a DRC check on my design, it comes up with hundreds of errors associated with curves. For example, I have the layer named Isolation in the shape of a rectangle with rounded corners; the curves have been designed such that the outer and inner curves are concentric. The DRC then needs to check that the width of the feature is not less than 10um all the way around (shown in Figure 1). After the DRC script is run, it finds errors around the curves as the curves are approximated with straight lines (shown Figure 2). How can I solve this problem?