Grounding in circuit simulations

[Weylin]

I notice that a lot of circuit simulations throw a fit if no ground terminal is found.
Why is this? Wouldn't a closed circuit with a two terminal voltage source and a proper load be enough?

Not all devices have a ground to connect to, such as a portable radio. What I do, and I'm unsure if this is accurate, is I put a very high value resistor between the ground and the rest of the circuit.
Does this properly simulate such a circuit and act as the interface between the circuit and surrounding air?

I'm really not sure what the point of a ground symbol is unless you're using one terminal voltage rails, modeling a 3 pronged outlet, or don't want to have a thousand vertical lines on your schematic plunging through everything to connect to the return line (or source line if you visualize electron flow )

 

[esp1]

hi W,
Its not really a circuit Ground, its Common Reference point for the simulation calculations and plots.

E

 

[Audioguru]

Ground is not an earth ground, it is just a wire that is common to many parts of a circuit.
0V in most circuits including a portable radio is the circuit's ground.

 

[BradtheRad]

A ground icon automatically sets 0V every place you put it.

Often it reduces clutter if you use ground icons.

Other times you may want instead to draw a line from one ground-connected component to another, to clarify there is a give-and-take relationship as to current flow.

 

[Weylin]

A ground icon automatically sets 0V every place you put it.

It seems redundant, since I would assume that all connections to the negative end of the voltage source terminal would be the 0 node.
Falstads simulator doesn't mind it except in a couple situations, but some simulators like to send positrons out the + side and into the aether, apparently :lol:

 

[godfreyl]

It seems redundant, since I would assume that all connections to the negative end of the voltage source terminal would be the 0 node.

That's a really bad assumption. What if you actually want to connect the positive terminal to ground? What if there's more than one voltage source? What if there's no DC source, only an AC source?

 

[FvM]

Falstads simulator doesn't mind it except in a couple situations, but some simulators like to send positrons out the + side and into the aether, apparently

Circuit simulators don't know about positrons. They are numerically solving network equations instead.

From an abstracted viewpoint, the solution shouldn't depend on superimposed common mode voltage, chosen zero volt reference point and grounding. But simulators are working with floating point numbers and have finite numeric accuracy. Circuit simulators are generally nonlinear solvers that have to use an iterative method. If node resistance against ground approaches infinity, convergence fails.