The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> Convergence of Spectre
https://designers-guide.org/forum/YaBB.pl?num=1138918410

Message started by milkdragon on Feb 2^nd, 2006, 2:22pm

Title: Convergence of Spectre
Post by milkdragon on Feb 2^nd, 2006, 2:22pm

Hi,

I have a question in PSS simulation. In the menu you can choose 3 different accuracy. Does it simply changing the reltol? Or can i say conservative and reltol=1e-4 gives higher accuracy than moderate and reltol=1e-5?

Thank you

Title: Re: Convergence of Spectre
Post by ACWWong on Feb 3^rd, 2006, 4:19pm

setting conservative does a bit more than just cahnging reltol. from what i remember it changes stuff like ensure high order is on (which makes the noise floor much lower), and changes other setting like lteratio, steady ratio and limits the timestep size etc. Anyway the accuracy/tolerable convergence error is a function of reltol and other things like vabstol, lteratio etc. For definitive detail, just look it up in the spectreRF reference manual.