You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right. As we all know, there's no such thing as free energy. Conservation of energy is a harsh mistress.
Which is why this simple circuit, which contains no power source but generates infinite voltage, seems like it has to be a bug.
In case the link breaks, this is literally just an LM324 "Op Amp (real)" with its supply voltage pins connected together:
I was trying to simulate this circuit (source), which relies on the current path through the IC to drain the caps when no voltage source is connected. Since that required a model for the inverter that includes simulation of the supply voltage, I was using the LM324 "Op Amp (real)". Suffice to say, the "phantom" voltage source in the LM324 model interfered with my simulation attempt.
The text was updated successfully, but these errors were encountered:
Right. As we all know, there's no such thing as free energy. Conservation of energy is a harsh mistress.
Which is why this simple circuit, which contains no power source but generates infinite voltage, seems like it has to be a bug.
In case the link breaks, this is literally just an LM324 "Op Amp (real)" with its supply voltage pins connected together:
I was trying to simulate this circuit (source), which relies on the current path through the IC to drain the caps when no voltage source is connected. Since that required a model for the inverter that includes simulation of the supply voltage, I was using the LM324 "Op Amp (real)". Suffice to say, the "phantom" voltage source in the LM324 model interfered with my simulation attempt.
The text was updated successfully, but these errors were encountered: