Parameters
==========

Design parameters are the ones the designer can tune in order to satisfy 
the design requirements. 

.. figure:: gui-project-parameters-design.png
	:scale: 75%
	
	Setting up the design parameters. 

For every design parameter you must specify its initial value. For design 
parameters that will be tuned by the optimizer you must also specify the 
lower and the upper bound. 

.. figure:: gui-project-parameters-operating.png
	:scale: 75%
	
	Setting up the operating parameters. 

Operating parameters define the environment in which the circuit operates. 
Typically a circuit must satisfy the design requirements across a range 
of operating parameter values. Therefore every operating parameters 
requires a nominal value which must lie within the interval specified by 
the lower and the upper bound. Circuit temperature is a special parameter 
in PyOPUS simulators. It is accessed as the ``temperature`` netlist 
parameter. 

.. figure:: gui-project-parameters-statistical.png
	:scale: 75%
	
	Setting up the statistical parameters. 

Statistical parameters are the ones that vary randomly. A designer has 
little or no influence on these parameters. Two statistical parameters are 
defined for every MOS transistor instance in file ``mosmm.inc``. For transistor
xmn1 the two parameters are accessible as mn1vt and mn1u0 netlist parameters
(they were generated by the netlister when the Miller opamp definition was 
netlisted). These parameters are supposed to be independent normally 
distributed random variables with zero mean and variance one. They model the 
random variations which give rise to the mismatch effect. 

A designer can reduce mismatch by increasing the transistor area which 
reduces the magnitude of random variations. If we want to simulate a circuit 
that has no random variations these parameters should be set to zero. 

Additionally four more statistical parameters are defined in the Monte Carlo 
MOS transistor model (gvtnmm, gu0nmm, gvtpmm, and gu0pmm). They are also 
supposed to be independent normally distributed random variables with zero 
mean and variance one. They represent the global process variations which 
affect all transistors of the same type in the same way. We list them among 
statistical parameters although they are not required by the netlist if we 
use a MOS transistor model other than ``mc`` (i.e. Monte Carlo model). 

Statistical parameters are currently not used in the GUI because none of the 
analyses that requires them is supported at this point. Nevertheless, they are 
added to the simulator's input netlist if they are defined. 

All parameter names must be unique identifiers.
