Organizing about "IC", "converge" "node set"...

I don't really understand what you're asking here. Perhaps you don't know what the difference between a nodeset and an initial condition is? Let me give a quick summary in the hope that this is what your issue is.

A nodeset is a means of helping DC convergence (and remember DC convergence is normally performed for most analyses; small-signal analyses need it to linearise the equations around that point; large-signal analyses such as tran need it to find the starting point for the analysis). What happens is that you specify a voltage on a node, and then that voltage is applied via a resistor. In fact here's the info from "spectre -h dc":

Nodesets help find the DC or the initial transient solution. You can specify
nodesets in the circuit description file with nodeset statements or in a
separate file by using the readns parameter. When nodesets are specified,
Spectre computes an initial guess of the solution by performing DC analysis,
while forcing the specified values on to nodes by using a voltage source in
series with a resistor whose resistance is rforce. Spectre then removes these
voltage sources and resistors and computes the required solution from this
initial guess.

Nodesets have two important uses. First, if a circuit has two or more
solutions, nodesets can bias the simulator towards computing the required
solution. Second, this is a convergence aid. By estimating the solution of the
largest possible number of nodes, you might be able to eliminate a convergence
problem or significantly speed up convergence.

The default rforce is 1Ohm, and is set via the options statement in spectre.

An initial condition is stronger. It use the same mechanism, of a voltage source in series with a resistor, but it is kept throughout the DC solution. So a nodeset drives the node until it converges, then removes the force and carries on converging - so the nodeset is a hint to pull the solution to the expected place, but there's no guarantee the force will persist because the solution may be nowhere near the forced value. With an initial condition the force is retained until it converges and then moves onto the subsequent analysis. So you would typically see the result as being close to the driven ic value (not identical because of the 1Ohm resistor, but close).

Normally initial conditions are honoured in time-domain analyses (such as tran) but ignored in dc and ac analyses (that can be altered with the ic parameter on tran, or force parameter on dc/ac). That's because it's dangerous to find a small-signal response with an initial condition because it may be a false situation; with transient at least the circuit can recover from the artificial starting point.

Anyway, I've written a lot here - and your issue may be nothing to do with the above!

Andrew

Hi Andrew,

Thanks for explain above.

(1) I got difference between IC and Nodeset.

      Node set: It if for initial guess with 10ohm and voltage source.

      IC: It is used during initial guess and dc-solution with resistor and voltage source.

(2) I had a misunderstanding about the nodeset. Nodeset was not effective to dc0p cause of I was forgetting to exist a forcing the node by desVar.

(3) What is the "converge('node_A "5.0")"? It is same to nodeset("node_A" 5.0) ?

Thanks,

Ichiro

Hi Ichiro,

Some answers to your points - first of all, the resistance for nodeset and ic is by default 1Ohm, not 10Ohms (it's controlled by the rforce option). It's the same - the difference is the duration of how long the "force" is held for (just during initial part of convergence, or for the enter convergence phase).

The converge function is just another way of defining a nodeset or initial condition:

converge('nodeset "node_A" 5.0) ; equivalent to nodeset("node_A" 5.0)
converge('ic "node_A" 5.0) ; equivalent to ic("node_A" 5.0)

The first argument to converge is the type of convergence aid used, not the name of the node.

Andrew

Hi Andrew,

Thank you so much, I got it.

Hi Andrew,

Thank you so much, I got it.