[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: gEDA-user: How to do transient analysis?

Daniel Nilsson wrote:

On Fri, Apr 08, 2005 at 05:34:14PM -0600, David Logan wrote:


What would be the best way to do transient analysis using geda? I have pspice instructions, which are easy enough. I just define a circuit like:

gnd -> IDC -> t_Close (t=0) -> R -> L -> gnd. Then I go into "Transient analysis" setup, set "Print Step" and "Final Time", and I get a graph of my capcitor voltage or inductor current.



I'm not sure I understand what your circuit is supposed to do, are you
sending a DC current through an inductor ? That should not produce
anyting interesting in simulation at all...


It actually does produce something interesting between t=0 and t=L/R, because current changes at a rate of If+(Ii-If)e^(-t/(L/R)), and the voltage changes at L*di/dt of that.

The transient analysis that I am doing now is showing that rate of change of capacitors or inductors over that (usually) short time frame.

In ngspice you should be able to do something like this to simulate an
inductor in transient analysis:

*Simulation for inductor circuit

I1 0 source pwl (0 0 10n 0 12n 1)
R1 source 1 10
L1 1 0 10nH

.tran 100ps 50ns

.end

Save that to a file filename.sp, run:

ngspice -b filename.sp -r filename.raw -o filename.out

Then load filename.raw into gwave.



I'll give that a shot. For simple circuits, this will probably be fine. However, when it gets into the multi-switch circuits, the pspice simulations have actual settable switches. Transient analysis takes into account switch locations at the specified time frames, and correctly calculates/graphs voltages and currents respectively.

So I presume there is no equivalent in geda?

I can show you a single switch example and a corresponding graph: http://www.loganshouse.net/problem
(Keep in mind that there can be multiple switches, open or closed by default at t=0, and switched at t values > 0. This simple example just doesn't happen show one.)

Thanks!
David Logan