Lumped-Circuit Variables, Reference Directions & Sign Conventions

A circuit diagram is not a picture of the electric field. It is a compact model that assigns one current to each branch and one potential to each node, then constrains those variables with element laws and conservation laws. The model is useful when propagation time, radiation, and field variation inside each connection are negligible at the time and length scales of interest.

The three variables that carry the model #

Current is signed charge flow through a chosen cross-section:

$$i(t)=\frac{dq(t)}{dt}.$$

The arrow on a schematic defines which physical direction counts as positive. A negative solution means the actual current points opposite the arrow; it does not mean the analysis failed.

Voltage is a potential difference, never an absolute property of one terminal:

$$v_{ab}(t)=\phi_a(t)-\phi_b(t).$$

Changing the reference node adds the same constant to every node potential and leaves every branch voltage unchanged. This gauge freedom is why a connected circuit with $n$ nodes has only $n-1$ independent node voltages.

Instantaneous power combines the two:

$$p(t)=v(t)i(t).$$

Energy over an interval is $w=\int p(t)\,dt$. Power is therefore the quantity that checks whether an element is absorbing energy, returning stored energy, or acting as a source.

Passive sign convention #

Choose the current reference arrow to enter the terminal marked positive for voltage. Then:

If the current arrow instead leaves the positive-voltage terminal, absorbed power is $p=-vi$. Both conventions work; mixing them inside one calculation does not.

Where the lumped model stops working #

An ideal wire is one node only while its voltage drop and propagation delay are negligible. At high frequency or over long conductors, the wire acquires distributed resistance, inductance, capacitance, propagation delay, and radiation. The correct response is not to repair KCL or KVL; it is to replace the wire by a richer model such as a transmission line.

The same boundary applies inside components. A resistor is modeled by a terminal constraint, not by resolving its microscopic field and carrier motion. Circuit theory succeeds by preserving the terminal behavior needed by the surrounding network.

This convention makes power balance mechanical. It also prevents a common category error: changing the reference direction after solving can make a correct negative result look positive while silently changing the variable being reported.