开云体育

Just an observation: there are a number of opamp models (for example) that bring a V- (negative supply) pin out of the subckt. The assumption being that the device can be used with V- grounded or with a negative voltage supply rail. It turns out that the model only works if V- is actually grounded and not with a negative supply. This is good example of poor design.

If you have a sub-schematic, or .subckt, and an assumed internal grounded connection that is brought out to the top level, it is helpful and convenient to use the COM net (with its own triangular symbol) within the .subckt instead of GND (0) triangle and to bring this out through the symbol. You can still ground this at the top level. Or not. If it's grounded, you can still plot the current flowing in the COM pin to ground.

There are some situations where it is very difficult not to use actual GND, even implicitly. The digital (A) devices are an example. They have a hidden "device common", so the outputs can source current, apparently from nowhere. I try to avoid anything where KCL (Kirchhoff's Current Law) is not observed.

--