2. For an ideal capacitor, the branch current is directly proportional to the derivative of the branch
voltage, namely ,
i(t)=C
dv(t)
dt
(2)
where the proportionality constant C is called the capacitance and is measured in far ads named in
honor of the English Chemist and Physicist Michael Faraday (1791-1867).
3. For an ideal inductor, the branch voltage is directly proportional to the derivative of the branch current,
namely,
v(t)=L
di(t)
dt
(3)
where the proportionality co ns tant L is called the inductance and is measured in henrys,namedafter
an American scientist Joseph Henry (1797-1878).
2.3 Nodal Analysis and Mesh Analysis
We have learned in previous sections that three sets of equations (KCL, KVL, and Ohm’s law) are sufficient
to perform circuit analysis for any circuits. To be more specific, in so lving a circuit that comprises B
branches (i.e., it has B cir cuit elements) and N nodes, we are solving for 2B variables, i.e., B branch
voltages and B branch currents. Accordingly, we need 2B independent equations to obtain unambiguous
solutions. For a circuit with B branches and N nodes, there will be N − 1independentKCLequations,
B − N +1 independent KVLequations,andB branch element equations. Thus there are, altogether, 2B
independent equations. However, since B is usually large (much larger than N ), solving 2B equations may
be an onerous task. The simpler alternatives are nodal analysis and mesh analysis.Bothmethodsarederived
from the aforementioned three sets of equations, however, they solve a smaller set of alternative algebraic
variables. The nodal analysis first defines node voltages for each node as the p otential difference between
the designated node and the reference (ground) node. Then, itemploysKCLtosetuptheso-callednode
equations, expressing every term in the equation in terms of node voltages. Thus each node equation is a
KCL equation, except that the unknown algebraic v ariables are node voltages. Once all node voltages are
solved (with the node equations), we can obtain all branch voltages and branch currents. In essence, instead
of directly solving for all branch voltages and branch currents, Nodal Analysis solves for node voltages and
then use those solutions to determine all branch voltages andbranchcurrents.Similarly,themeshanalysis
defines mesh currents for each mesh (a closed loop) as the current traversing around this closed loop. Then,
it employs KVL to set up the so-called mesh equations , expressing wherever possible every term in the
equation in terms of mesh currents. Thus each mesh equation isaKVLequation,exceptthattheunknown
algebraic variables are mesh currents. Once all mesh currents are solved (with the mesh equations), we can
obtain all branch voltages and branch currents. Since the number of nodes (or that of meshes) is significantly
less than the number of branches, those two methods involve many fewer variables to solve for. All circuits
can be solved either by nodal analysis or by mesh analysis.
To facilitate the employment of no dal analysis and mesh analysis, we need to introduce another sign conven-
tion, i.e., the so-called associated passive sign convention which is defined in accordance with the physical
properties of passive circuit elements, i.e., circuit elements like resistors that absorb energy. The associated
passive sign convention states that the branch current of a passive circuit element flows from high to low
potential. Accordingly, the sign of the branch current is positive if it flows from the positive (+) polarity of
the v oltage to the negative (-) polarit y. As such, unless otherwise specified, once the polarity of the voltage
is defined, the direction of associated branch current is defined. The converse is also true. In other words,
if the current flow direction is marked, then the voltage polarity is defined accordingly.
The following section presents some e x amples that use nodal analysis and mesh analysis for simple resistive
circuits. The mathematical tools involved are simply that ofsystemsoflinearequations. Thoseequations
can be solved by hand (if the number of unknowns is less than or equal to 3), using Cramer’s rule, or using
acomputerprogramlikeMatlab.
4