User Guide

Worst-case analysis
307
Inputs
In addition to the circuit description, you need to provide
two pieces of information:
• the parameter tolerances
• a definition of what worst means
You can set tolerances on any number of the parameters
that characterize a model.
The criterion for determining the worst values for the
relevant model parameters is defined in the .WC
statement as a function of any standard output variable in
a specified range of the sweep.
In a given range, reduce the measurement to a single value
by one of these five collating functions:
Y ou can define Worst as the highest (HI) or lowest (LO)
possible collating function relative to the nominal run.
Procedure
To establish the initial value of the collating function,
worst-case analysis begins with a nominal run using all
model parameters at their nominal values.
Next, multiple sensitivity analyses determine the
individual effect of each model parameter on the collating
function. This is accomplished by varying model
parameters, one at a time, in consecutive simulations. The
MAX Maximum output variable value
MIN Minimum output variable value
YMAX Output variable value at the point
where it differs the most with the
nominal run
RISE_EDGE
(value)
Sweep value where the output
variable value crosses above a given
threshold value
FALL_EDGE
(value)
Sweep value where the output
variable value crosses below a given
threshold value
You can
d
e
f
ine mo
d
e
l
s
f
or near
l
y a
ll
primitive analog circuit parts, such as
resistors, capacitors, inductors, and
semiconductor devices. PSpice reads the
standard model parameter tolerance
syntax specified in the .MODEL statement.
For each model parameter, PSpice uses the
nominal, minimum, and maximum
probable values, and the DEV and/or LOT
specifiers; the probability distribution type
(such as UNIFORM or GAUSS) is ignored.
You can use ana
l
og
b
e
h
aviora
l
mo
d
e
l
s to
measure waveform characteristics other
than those detected by the available
collating functions, such as rise time or
slope. You can also use analog behavioral
models to incorporate several voltages and
currents into one output variable to which a
collating function may be applied. See
Chapter 6,
Analog behavioral
modeling, for more information.
T
h
is proce
d
ure saves time
b
y per
f
orming
the minimum number of simulations
required to make an educated guess at the
parameter values that produce the worst
results. It also has some limitations, which
are described in the following sections.
Pspug.book Page 307 Wednesday, November 11, 1998 1:14 PM