Path: news.mathworks.com!not-for-mail
From: "Antonin Ancelle" <antonin.ancelle@mathworks.com>
Newsgroups: comp.soft-sys.matlab
Subject: Re: C2000 DSP and Simulink : using a memory-stored variable as a block parameter
Date: Thu, 6 Mar 2008 14:18:36 +0100
Organization: The MathWorks, Inc.
Lines: 43
Message-ID: <fqoqve$498$1@fred.mathworks.com>
References: <fqogav$eol$1@fred.mathworks.com>
NNTP-Posting-Host: fr-ancellea.mathworks.fr
X-Trace: fred.mathworks.com 1204809518 4392 172.16.45.248 (6 Mar 2008 13:18:38 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Thu, 6 Mar 2008 13:18:38 +0000 (UTC)
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2900.3138
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3198
X-RFC2646: Format=Flowed; Original
Xref: news.mathworks.com comp.soft-sys.matlab:455718



Hi Kevin,

I am going to send you a direct email with a burst model of a PI controller 
where you can dynamically change the parameters.
It has no D component, but you can easily add it if you have a need for it.
It is more conventional than the TI version in the sense that the P 
component is not cascaded with the I.

On the TI version if you put P = 0, the rest will be 0 regardless of what I 
or D component you have.

I will let you play with it.

Good luck,

Antonin.


"Kevin " <kedecuyp@mathworks.com> wrote in message 
news:fqogav$eol$1@fred.mathworks.com...
> Hello,
>
> I've been toying with the C2000 toolbox for a little while
> now, but I'm hitting a wall on this one.
>
> What I'm trying to do is a PID controler whose parameters
> can be changed by an external source or using a RTDX channel.
> In fact this problem might arise for every block whose
> parameters can't be specified from the outside.
>
> Now I could create the controler with basic Simulink blocks,
> but I want to use to more efficient PID-Controler from the
> DMC library.
>
> Since the C2000 doesn't have a workspace, I tried to use the
> "Memory Copy" block from the Link for Code Composer Library
> with no results since the parameters must come from the
> workspace.
>
> Is there a way to tell the block the the value of a
> parameters must be taken from memory and not from the workspace?