A few historically relevant contributions to solving network equations, and some related physical/mechanical excercises
Keep in mind when "simulating" linear, time-invariant electrical networks, that it can be proven that the equations coming from the MNA method are the actual differential equations governing the network behavior, no approximation. And keep in mind that phrasing the Kirchhoff basics in matrix form, under very broadly applicable conditions can be proven to lead to a formally solvable system of equations, for every possible network topology. Why iterate like a sort of eigenvalue search over the equations, when you have no idea of the conditioning of your state progression operator, when the formal system of equation can inverted without iterations ? Moreover the simulated systems are not variable so the simulation is done on a fixed network, and their are fairly standard re-conditioning matrix solutions available, either for a numeric (inversion and back-substitution !) or a formal s-domain transformed solution. Just doing an iteration without a clou if it is convergent, well-conditioned or stable should at least be well tested against a formal solution, and there might be nice (first year EE) Linear Algebra transformation to improve the (also important) numerical stability of your iteration progress. T. -- dupswapdrop -- the music-dsp mailing list and website: subscription info, FAQ, source code archive, list archive, book reviews, dsp links http://music.columbia.edu/cmc/music-dsp http://music.columbia.edu/mailman/listinfo/music-dsp
