Date: Oct 23, 2012 9:46 PM
Author: lady firmament
Subject: Re: error in integrator

dpb <none@non.net> wrote in message <k64g9e$tnn$1@speranza.aioe.org>...
> On 10/22/2012 5:05 PM, Phil Goddard wrote:
> > You need to figure why the signal being fed into the integrator is
> > becoming nan or inf.
> >
> > There can be many reasons for that happening and nobody is going to be
> > able to tell you specifically why your model is doing it without seeing
> > the model.
> > However, you most likely have a divide by zero somewhere.

>
> Which, of course (to help OP in the looking for the cause of which 'cuz
> I know you know this) doesn't have to be and likely isn't an actual zero
> in a denominator but a term that vanishes over time as, perhaps, some
> other term grows. Exponentials and their ilk are prime candidates as
> would/could be a trig function going through a null point, etc., etc.,
> etc., ...
>
> --

Thanks you for valuable suggestions. Can I post the model in this page or not?
Let me know!
Firmament