User Guide

8
-
2 Tutorial: Fittin
g
Model Data (Bipolar Transistor)
Usin
g
the PSpice
Optimizer to Fit Data to
Model Parameters
The PSpice Optimizer can fit a parameterized model to one or
more sets of data points. The source for this external data might
be:
• Results from measuring a real device (e.g., using a
semiconductor curve tracer).
• A manufacturer’s data sheet.
• A PSpice simulation.
To fit a model using the PSpice Optimizer, you need a text file
containing the measured values and the measurement points.
Example: For static I-V characteristic of a diode, the external
data file would contain pairs of values where each pair consists
of the voltage at which the measurement was made and the
current through the device at that voltage.
The PSpice Optimizer performs a least squares estimation of the
parameter values that result in the best fit between the external
data and the values produced by the model. More formally,
given a set of N data points {x
i
,y
i
}, a model that has M
parameters {a
j
}, and a model relationship of the form:
,
the optimizer chooses the parameters {a
j
} to minimize:
yx
()
yxa
1
…
a
M
;
()
=
y
i
y
–
x
i
a
1
…
a
M
;
()[]
2
i1=
N
∑