Is nodal analysis should be applied only when circuit containing current sources..? Or otherwise we have to use source transformation.. When any voltage source is present in the circuit..?

  • \$\begingroup\$ Nodal Analysis can be applied when voltage sources are present, and indeed it is often best suited to this situation. \$\endgroup\$
    – jramsay42
    Commented Nov 29, 2017 at 11:10
  • 1
    \$\begingroup\$ Nodal Analysis applies to all circuits. It is independent on the components used be they sources or not. You can do nodal Analysis on a resist-only network if you like. \$\endgroup\$ Commented Nov 29, 2017 at 11:12
  • \$\begingroup\$ @Bimpelrekkie, nodal analysis per se can't be used in the presence of voltage sources. Modified nodal analysis is used to accommodate voltage sources. \$\endgroup\$
    – The Photon
    Commented Nov 29, 2017 at 15:55
  • \$\begingroup\$ OP, (Modified) nodal analysis is taught mainly because it's the algorithm used in SPICE and other circuit simulation programs. For pen & pencil analysis we usually use ad hoc methods, and only fall back on NA when a circuit is so messy we should probably just use SPICE anyway. \$\endgroup\$
    – The Photon
    Commented Nov 29, 2017 at 15:59
  • \$\begingroup\$ @ThePhoton No doubt you're right, I meant "Nodal Analysis" in the sense of solving circuits having nodes. I know how to do that but never bothered to remember the exact names of each method. My bad for calling all these "Nodal analysis" ;-) \$\endgroup\$ Commented Nov 29, 2017 at 16:01

2 Answers 2


I will try to explain in very simple terms.

Basically there are 2 methods Nodal Analysis and Mesh analysis.

Nodal Analysis is done based on Voltages/Potential present at different nodes in the network. If you want to represent a node in a circuit, you can represent that only with a potential i.e voltage, a current flows between nodes only when there is a potential difference between them. In nodal analysis the basic concept used is that total incoming current to a node is equal to total outgoing current, but to represent this incoming and outgoing current we use potential at the nodes (Current flowing through a resistor is equal to potential difference across resistor divided by the value of the resistor, if we consider 2 ends of a resistor as 2 nodes and V1, V2 be the potential at these 2 nodes, then potential/voltage difference across them will be V1-V2 and the current flowing through will be (V1-V2)/R)

In Mesh analysis, we use the concept that total voltage drop across a close circuit is zero, and we use current to compute drop across the different elements of the circuit (Note: current across a series circuit is same for all elements in the circuit)

I hope this will make things a little clear


As strictly defined, nodal analysis is not able to solve a circuit containing an independent voltage source (or any other element that does not have an "admittance representation"), because it depends on each branch current being able to be expressed as a function of the node voltages.

There is an extended version called modified nodal analysis that is able to solve circuits with voltage sources. Rather than use source transformations, this method uses the supernode concept to eliminate a problematic branch from the nodal analysis and introduce its effect through a supernode equation.

(Modified) nodal analysis is studied mainly because it's the method used in SPICE and other circuit simulators. Of course it's also good to have a well-defined method for solving arbitrary circuits by pencil & paper, but normally we us ad hoc methods for those problems to minimize the amount of algebra required.

  • \$\begingroup\$ It's all just currents in and currents out. For hand work, one can merely assign a variable to the unknown current of the voltage source. Given that you can then add an equation illustrating the relationship of the two node voltages, you'll still wind up with exactly N equations and N unknowns. All this without "supernode" thinking, though of course that works as well. \$\endgroup\$
    – jonk
    Commented Nov 29, 2017 at 17:47
  • \$\begingroup\$ @jonk, yes you can do that. But then it wouldn't be nodal analysis (as strictly defined) any more. \$\endgroup\$
    – The Photon
    Commented Nov 29, 2017 at 17:49
  • \$\begingroup\$ :) I'm not arguing about that. That is a matter for electronics engineers to argue over, themselves. It's just defining terms for accurate communications. And I'm just a hobbyist. However, I think in pictures -- like a painting of sorts -- and to me these are all of the same piece of cloth. No meaningful difference, at all. I'd see these alternatives, regardless of the terms applied. I'd still want to use the right terms when communicating and I have no right to suggest anything different to those whose professional life is here. I was just mentioning this point in case it helps the OP. \$\endgroup\$
    – jonk
    Commented Nov 29, 2017 at 18:16
  • \$\begingroup\$ @jonk, and I also realize that we don't have to use the strictest and most careful language all the time to communicate, which is why I said in my answer I was talking about nodal analysis "as strictly defined". \$\endgroup\$
    – The Photon
    Commented Nov 29, 2017 at 18:53
  • \$\begingroup\$ Understood. Then hopefully this discussion here doesn't confuse, but rather at least causes no harm and possibly helps the OP (or others reading along) a little as they consider arriving information. \$\endgroup\$
    – jonk
    Commented Nov 29, 2017 at 22:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.