help-gnucap
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Help-gnucap] Gnu cap src help


From: al davis
Subject: Re: [Help-gnucap] Gnu cap src help
Date: Sat, 25 Aug 2012 01:37:40 -0400
User-agent: KMail/1.13.5 (Linux/2.6.32-5-amd64; KDE/4.4.5; x86_64; ; )

On Friday 24 August 2012, Felix Salfelder wrote:
> is there any reason why you chose gnucap? gnucap does a lot
> to do fast numerical simulation of electrical networks,
> which is hardly needed for network reduction. also, a
> resistor network can be easily represented by a triangular
> incidence matrix over nonnegative numbers. using the gnucap
> data structures looks like overkill...
> 
> but of course i might be missing something, i'm just curious.

Why not do it on gnucap?

Adding the capability of an approximate solver for large passive 
networks would be most welcome, and consistent with gnucap's 
mission.

The way to do it is as a plugin alternative to the subckt, so 
you can use the new reduction based approximate solver along 
with the rest of gnucap ..  another kind of "mixed mode".  You 
could even add functions that check the validity of the reduced 
model and automatically use the full model when accuracy is 
needed or the reduced model when speed is more important than 
accuracy.

"precalc" "expand" "tr_begin" are places where the reduction 
code might go, so a user could use the reduced model easily 
along with other parts of the circuit non-reduced.

For those who don't know what we are talking about....
Why would you want this?  Answer ..  Full simulation with 
parasitics and interconnect.  Interconnect models often come 
from some kind of extractor, and are usually huge RC or RLC 
circuits with just a few ports.  Think of a transmission line 
modeled as 100000 RLC sections.

Real answers to this get a bit heavy.  I like to keep the "help" 
list beginner-friendly.  How about joining us on "gnucap-devel" 
and going into detail there?

al.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]