Designer’s Guide Community

‹ Previous topic | Next topic ›

Pages: 1

point to point ringing in some integration methods (Read 2425 times)

jefkat

Community Member

Offline

Posts: 69

point to point ringing in some integration methods
Apr 22^nd, 2007, 8:54am

Hi..
I am trying to understand why would some integration methods result in ringing (especially when the time constants are smaller than the time step). Intutively thinking I would miss some signal detail (due to fast poles) but downright ringing !! cant seem to make sense of it..Anyone with some commenst to put me on the right track of thinking?

Thanks much!
jefkat

IP Logged

mg777

Senior Member

Offline

Posts: 131

Re: point to point ringing in some integration met
Reply #1 - Apr 22^nd, 2007, 11:37pm

Think of an integration algorithm as a difference equation, with an impulse response. You can have underdamped filters like Adams-Bashford which use extrapolation. Implicit methods can reduce the overshoot in the impulse response, while Runge-Kutta will tend to be damped.

Refer to the revered 'Numerical Methods in C' for further illumination.

M.G.Rajan
www.eecalc.com

IP Logged

Andrew Beckett Senior Fellow Offline Life, don't talk to me about Life... Posts: 1742 Bracknell, UK	Re: point to point ringing in some integration met Reply #2 - Apr 30^th, 2007, 5:37am Or for a more accessible book for designers, look at Ken's book http://www.designers-guide.org/Books/#Kundert-1995 A Designer's Guide to SPICE and Spectre, page 150. Regards, Andrew.
Back to top	IP Logged

Pages: 1

‹ Previous topic | Next topic ›