SV: Re: SV: Re: SV: Re: [sdiy] Simulating SW?

[harrybissell]

I'd like to introduce a point in favor of "less good models".

In some cases, a model can have TOO much detail that can cause it to
actually hurt the simulation.  I was doing a three-phase (SCR-diode)
bridge rectifier circuit.  The "good" SCR model was causing the
"one nano-second... TWO nano-second..." response typical of a simulation
that has some "issues".

I replaced the GOOD SCR model with a current controlled switch and two diodes.
One diode adds gate gurrent into the switch control, the second causes the latch.
It has a one diode forward drop, and shuts off when the holding current is gone.
ISN'T that an SCR ???

The new model ran very well.  I'll add that I was testing another part of the
circuit,
and that I already KNEW the SCR part works in real life.

The idea is... does the simulation reflect real life 'close enough' to give a
useful
result ?  I'd say you will never know, unless you both simulate AND build...

H^) harry