Star-Hspice has four voltage and current controlled elements, known as E, G, H, and F Elements. You can use these controlled elements in Star-Hspice to model both MOS and bipolar transistors, tunnel diodes, SCRs, as well as analog functions such as operational amplifiers, summers, comparators, voltage controlled oscillators, modulators, and switched capacitor circuits. The controlled elements are either linear or nonlinear functions of controlling node voltages or branch currents, depending on whether you use the polynomial or piecewise linear functions. Each controlled element has different functions:
The following sections discuss the polynomial and piecewise linear functions and describe element statements for linear or nonlinear functions.
The controlled element statement allows the definition of the controlled output variable (current, resistance, or voltage) as a polynomial function of one or more voltages or branch currents. You can select three polynomial equations through the POLY(NDIM) parameter.
The POLY(1) polynomial equation specifies a polynomial equation as a function of one controlling variable, POLY(2) as a function of two controlling variables, and POLY(3) as a function of three controlling variables.
Along with each polynomial equation are polynomial coefficient parameters (P0, P1 ... Pn) that can be set to explicitly define the equation.
If the function is one-dimensional (a function of one branch current or node voltage), the function value FV is determined by the following expression:
The following controlled source statement is an example of a one-dimensional function:
The above voltage-controlled voltage source is connected to nodes 5 and 0. The single-dimension polynomial function parameter, POLY(1), informs Star-Hspice that E1 is a function of the difference of one nodal voltage pair, in this case, the voltage difference between nodes 3 and 2, hence FA=V(3,2). The dependent source statement then specifies that P0=1 and P1=2.5. From the one-dimensional polynomial equation above, the defining equation for V(5,0) is
Where the function is two-dimensional (a function of two node voltages or two branch currents), FV is determined by the following expression:
For a two-dimensional polynomial, the controlled source is a function of two nodal voltages or currents. To specify a two-dimensional polynomial, set POLY(2) in the controlled source statement.
For example, generate a voltage controlled source that gives the controlled voltage, V(1,0), as:
To implement this function, use the following controlled source element statement:
This specifies a controlled voltage source connected between nodes 1 and 0 that is controlled by two differential voltages: the voltage difference between nodes 3 and 2 and the voltage difference between nodes 7 and 6, that is, FA=V(3,2) and FB=V(7,6). The polynomial coefficients are P0=0, P1=3, P2=0, P3=0, P4=0, and P5=4.
For a three-dimensional polynomial function with arguments FA, FB, and FC, the function value FV is determined by the following expression:
For example, generate a voltage controlled source that gives the voltage as:
from the above defining equation and the three-dimensional polynomial equation:
Substituting these values into the voltage controlled voltage source statement yields the following:
The above specifies a controlled voltage source connected between nodes 1 and 0 that is controlled by three differential voltages: the voltage difference between nodes 3 and 2, the voltage difference between nodes 7 and 6, and the voltage difference between nodes 9 and 8, that is, FA=V(3,2), FB=V(7,6), and FC=V(9,8). The statement gives the polynomial coefficients as P1=3, P7=4, P19=5, and the rest are zero.
The one-dimensional piecewise linear function allows you to model some special element characteristics, such as those of tunnel diodes, silicon-controlled rectifiers, and diode breakdown regions. The piecewise linear function can be described by specifying measured data points. Although the device characteristic is described by some data points, Star-Hspice automatically smooths the corners to ensure derivative continuity and, as a result, better convergence.
A parameter DELTA is provided to control the curvature of the characteristic at the corners. The smaller the DELTA, the sharper the corners are. The maximum DELTA is limited to half of the smallest breakpoint distance. If the breakpoints are quite separated, specify the DELTA to a proper value. You can specify up to 100 point pairs. At least two point pairs (four coefficients) must be specified.
In order to model bidirectional switch or transfer gates, the functions NPWL and PPWL are provided for G Elements. The NPWL and PPWL function like NMOS and PMOS transistors.
The piecewise linear function also models multi-input AND, NAND,OR, and NOR gates. In this case, only one input determines the state of the output. In AND / NAND gates, the input with the smallest value is used in the piecewise linear function to determine the corresponding output of the gates. In the OR / NOR gates, the input with the largest value is used to determine the corresponding output of the gates.
Star-Hspice Manual - Release 2001.2 - June 2001
