Description
===========

Bug 372 was an enhancement request, asking for a PSPICE-compatible instance
parameter M that simulated a number of parallel MOSFETS:
   Mfoo 1 2 3 4 MODEL M=2
should be the same as
   Mfoo 1 2 3 4 MODEL
   Mbar 1 2 3 4 MODEL

The tests in this directory verify that  the M= parameter does indeed produce
results that are the same as the equivalent parallel MOSFETS.

Verification
============

For each level of mosfet there are two netlists, one with parallel mosfets,
the other with a single mosfet with M=2.  Run all the netlists, then make
the comparisons as so:

   xyce_verify.pl invert1.cir invert1.cir.prn invert1-sim.cir.prn invert1.plot
   xyce_verify.pl invert3.cir invert3.cir.prn invert3-sim.cir.prn invert3.plot
   xyce_verify.pl invert_bsim3.cir invert_bsim3.cir.prn \
                  invert_bsim3-sim.cir.prn invert_bsim3.plot
   xyce_verify.pl invert_b3soi.cir invert_b3soi.cir.prn \
                  invert_b3soi-sim.cir.prn invert_b3soi.plot

Ideally all four xyce_verify runs should produce no warnings or comparison
failures.  The tolerance of the comparisons have been set in the netlists
so that this would be the case for Release 2.0.  These are netlists with
square-wave driving voltages, though, so comparison failures would not indicate
real testing failures.  If the code flags signals as having too large an error,
the plot files should be analysed by hand -- the signals for the invert1.cir.prn
output should match those of the invert1-sim.cir.prn, with significant relative
errors only being seen at the transitions between high and low --- those
spikey differences can be ignored, and the real test is that the high values
should line up exactly.

NOTE ADDED 13 July 2005

The invert_b3soi circuits will not run in versions of Xyce prior to
3.0.  Versions of Xyce prior to Release 2.1 had no level 10 mosfet,
and Release 2.1 had a version of the BSIM SOI model that was not
complete.
