Electrical Engineering › Section E · Methods of analysis (dc) › Topic E.4

Nodal analysis

Nodal analysis makes the voltages at a circuit's nodes the unknowns and writes KCL at each one. It usually needs the fewest equations of any method, it works for every circuit, planar or not, and it is the basis of the simulators engineers use every day.

Voltages as the unknowns

"Mesh analysis" chose loop currents so that KCL took care of itself. Nodal analysis makes the opposite choice. Pick one node as the reference, ground, at 0 V, and let the voltage of every other node relative to it be an unknown. Once the node voltages are known, every branch current follows from Ohm's law, and every element's voltage is the difference of two node voltages, so KVL takes care of itself. What remains is KCL at each non-reference node: a circuit with \(n\) nodes needs only \(n - 1\) equations, fewer still if voltage sources fix some node voltages directly. Nodal analysis also works on non-planar circuits, where mesh analysis cannot.

Reference node and branch currents

Any node can be the reference, but the one with the most branches, often the − terminal of the main source, gives the simplest equations. All other voltages are then measured from it, exactly as you would probe a circuit with the black lead on ground. The key relation is the current in a resistor \(R\) between nodes a and b:

I_{a \to b} = \frac{V_a - V_b}{R}

If b is the reference, this is just \(V_a / R\). A current flowing the other way simply comes out negative, so you never have to guess directions correctly; you only have to be consistent within each node's equation.

The general approach

Step 1: choose the reference node and label the others \(V_1\), \(V_2\) and so on. Step 2: at each non-reference node, write KCL. A tidy convention is to treat every resistor current as leaving the node, \((V_{\text{node}} - V_{\text{other}})/R\), and set the sum of currents leaving equal to the sum of source currents entering. Step 3: substitute any node voltages fixed by voltage sources connected to ground. Step 4: solve the equations together. Step 5: find any branch current or element voltage from the node voltages. Because each equation is written from its own node, the same resistor appears in two equations with opposite signs, which is a useful check.

Worked example: the three-mesh circuit again

Return to the circuit of three windows in "Mesh analysis": the 15 V source, + at the top, feeds node a through 1.0 kΩ; 2.2 kΩ runs from a to ground; 1.0 kΩ from a to b; 3.3 kΩ from b to ground; 1.5 kΩ from b to node c; and the 6 V source, + at c, holds c at 6 V. Work in kΩ, mA and V. 1. The source nodes are fixed at 15 V and 6 V, leaving only \(V_a\) and \(V_b\). 2. Node a: \(\frac{V_a - 15}{1.0} + \frac{V_a}{2.2} + \frac{V_a - V_b}{1.0} = 0\). 3. Node b: \(\frac{V_b - V_a}{1.0} + \frac{V_b}{3.3} + \frac{V_b - 6}{1.5} = 0\). 4. Solving: \(V_a = 8.75\,\text{V}\) and \(V_b = 6.47\,\text{V}\), as mesh analysis found, but with two equations instead of three.

The format approach

For circuits with resistors and independent current sources only, the equations can be written by inspection. Convert any practical voltage sources first, as in "Current sources and source conversions". In the equation for node \(k\), the coefficient of \(V_k\) is the sum of the conductances connected to node \(k\); the coefficient of each neighbouring node voltage is minus the conductance joining the two; and the right side is the sum of source currents entering node \(k\), negative for those leaving. In matrix form this is

\mathbf{G}\mathbf{V} = \mathbf{I}

where \(\mathbf{G}\), the conductance matrix, is symmetric about its diagonal for circuits like these. Conductances in millisiemens pair neatly with currents in milliamperes and voltages in volts.

Worked example: two current sources

A 9 mA current source pushes current from ground into node 1, and a 1 mA source draws current from node 2 to ground. \(R_1 = 2.0\,\text{k}\Omega\) runs from node 1 to ground, \(R_2 = 2.0\,\text{k}\Omega\) from node 1 to node 2 and \(R_3 = 3.0\,\text{k}\Omega\) from node 2 to ground. 1. Conductances: 0.5, 0.5 and 0.333 mS. 2. Node 1: \(1.0V_1 - 0.5V_2 = 9\). 3. Node 2: \(-0.5V_1 + 0.833V_2 = -1\). 4. Solving: \(V_1 = 12\,\text{V}\) and \(V_2 = 6\,\text{V}\). 5. Currents: 6 mA down \(R_1\), \((12 - 6)/2.0 = 3\,\text{mA}\) through \(R_2\) and 2 mA down \(R_3\). Check KCL at node 2: 3 mA in, \(2 + 1 = 3\,\text{mA}\) out.

Voltage sources and supernodes

A voltage source from a node to the reference fixes that node's voltage, removing an unknown, as in the first worked example. A voltage source between two non-reference nodes is harder, because its current is unknown and cannot be written with Ohm's law. Treat the two nodes and the source as one region, a supernode. Write one KCL equation for the whole region, adding every current that crosses its boundary but ignoring the current inside the source. Then add the constraint that the difference of the two node voltages equals the source voltage. The pair replaces the two ordinary node equations, mirroring the supermesh of mesh analysis.

Worked example: a supernode

An 8 mA current source pushes current from ground into node 1. \(R_1 = 2.0\,\text{k}\Omega\) runs from node 1 to ground and \(R_2 = 3.0\,\text{k}\Omega\) from node 2 to ground. A 6 V source connects the two nodes, + at node 1. 1. Constraint: \(V_1 - V_2 = 6\). 2. Supernode KCL, current in equals current out: \(8 = \frac{V_1}{2.0} + \frac{V_2}{3.0}\). 3. Substitute \(V_1 = V_2 + 6\): \(8 = 0.5V_2 + 3 + 0.333V_2\), so \(V_2 = 6\,\text{V}\) and \(V_1 = 12\,\text{V}\). 4. \(R_1\) carries 6 mA and \(R_2\) 2 mA, which add to 8 mA. 5. KCL at node 1 alone shows that 2 mA flows from node 1 through the voltage source to node 2.

Solving and simulating

Two or three node equations can be solved by elimination or Cramer's rule; larger systems call for matrix methods, a calculator or software. Circuit simulators such as SPICE, developed at the University of California, Berkeley, in the early 1970s, build their equations with an extended form called modified nodal analysis, which adds the currents of voltage sources as extra unknowns. Some circuits contain controlled (dependent) sources, whose value depends on a voltage or current elsewhere, as in transistor models. Nodal analysis handles them with one extra equation expressing the controlling quantity in node voltages, although the matrix is then no longer symmetric.

Common mistakes to avoid

First, mixing conventions within one node equation, treating some currents as leaving and others as entering without matching signs. Second, writing a KCL equation at the reference node, which adds nothing. Third, trying to give a voltage source's current an Ohm's-law expression; it has none, so use a fixed node voltage or a supernode. Fourth, in the format approach, putting resistances rather than conductances in the matrix, or forgetting to convert voltage sources first. Fifth, counting a current source's internal parallel resistance twice after a conversion.

Key takeaways

Nodal analysis takes the \(n - 1\) non-reference node voltages as unknowns and writes KCL at each node, with each resistor current \((V_a - V_b)/R\). Choose the most connected node as the reference. A voltage source to ground fixes a node voltage; one between two nodes forms a supernode with a constraint equation. By inspection, \(\mathbf{G}\mathbf{V} = \mathbf{I}\): diagonal entries are total conductances at each node, off-diagonals minus shared conductances, and the right side the source currents entering. SPICE uses modified nodal analysis.

Flashcards

Flashcard

Nodal analysis

Node voltages as unknowns, KCL at each non-reference node

KVL is automatic.

Test your understanding

Question 1 of 5

A circuit has 6 nodes and no voltage sources. How many nodal equations are needed?

Question 2 of 5

Node a is at 9 V and node b is at 4 V. What current flows from a to b through a 2.5 kΩ resistor?

Question 3 of 5

A 10 V source (− at ground) feeds node a through 1.0 kΩ, and two 2.0 kΩ resistors run from a to ground. What is Va?

Question 4 of 5

Using the format approach, what is the diagonal coefficient for a node joined to ground by 2.0 kΩ and to another node by 5.0 kΩ?

Question 5 of 5

A 4 V source joins node 1 (+) and node 2. Node 1 has 1.0 kΩ to ground, node 2 has 1.0 kΩ to ground, and a 6 mA current source feeds node 1. What is V2?

← Previous topicNext topic →