Definition of electrical branch

[nebbione]

Hi everyone! I wanted to know why in the circuit analysis a generator doesn't represent an electrical branch?
And the second question is if two resistors are in series on a wire, does it represents only a branch(the series of the resistance) or two branches ?

 

[Studiot]

A branch is a single length conductor with zero impedance connecting two nodes.
The potential is the same along the whole length of the branch and at both nodes.
There is therefore only one current flowing in any branch; this is called the branch current.

Does a generator conform to this definition?

 

[nasu]

A branch is a single length conductor with zero impedance connecting two nodes.

Zero impedance? Where did you get this requirement from?
I did not find anywhere (yet) this definition. I thought that anything between two nodes is a branch (if there is no other node).

Just a couple of internet places where they don't assume this zero impedance for a branch:
http://en.wikipedia.org/wiki/Nodal_analysis
http://www.allaboutcircuits.com/vol_1/chpt_10/2.html

Regarding OP: Can you show details of the situation where the generator is not considered "a branch"?

 

[Studiot]

Strictly, even a resistor has a node at each end.

You connect nodes by branches.

Mostly humans are intelligent enough to condense this to including the resistor in the 'branch', but computers are not.

Thus a generator is simply a 'component' with two nodes (terminals).

 

[nasu]

Strictly, even a resistor has a node at each end.

Even when you have two series resistors in a "branch"?

I thought that to have a node you need to have at least three branches meeting at that point. Not every point or terminal is a node. In a simple series circuit with 10 resistors and one source, for example, there are no nodes.

PS. Maybe your terms are from a different field than circuit analysis based on Kirchhoff's rules?

 

[nebbione]

So a resistor is one branch and two resistors in series still one branch ?

 

[Studiot]

Even when you have two series resistors in a "branch"?

In electrical networks almost every component has to have at least two connection points to the network. The main exception that can make do with a single connection is Earth or ground.

When you are programming a computer, you have to tell the computer where these connection points are. This specifies the network to the computer.

Each connection point has to have a name or identifier.

I am calling these connection points nodes and the connections branches.

Normally you specify by what is known as the incidence matrix or a netlist.

The specification of other non electrical networks, eg pipes or roads or structural framworks follows the same pattern, except that some networks can function with only one connection to some components.

 

[nasu]

Each connection point has to have a name or identifier.
I am calling these connection points nodes and the connections branches.
Normally you specify by what is known as the incidence matrix or a netlist.

OK, so we were talking about two different things. The "classical" Kirkchoff's rule "nodes" are a little different. I suppose it is left to OP to specify what is his actual problem.

 

[Studiot]

This thread is a spin off from another about this by nebbione, where he clearly indicates he wants to use matrix methods.
Classic Kirchoff is not a matrix method.

https://www.physicsforums.com/showthread.php?t=641568

As a matter of interest, how would you define the branch currents in Maxwell's mesh method?

 

[nasu]

This thread is a spin off from another about this by nebbione, where he clearly indicates he wants to use matrix methods.
Classic Kirchoff is not a matrix method.

OK, I had no idea about the other thread. Sorry for my interventions.