Fwd: Q'slew behavior with a discontinuous Q

---------- Forwarded message ----------
From: ZHICHAO DENG <zhichaodeng@gmail.com>
Date: Wed, Sep 12, 2012 at 7:43 PM
Subject: Re: Q'slew behavior with a discontinuous Q
To: Joachim Haase <Joachim.Haase@eas.iis.fraunhofer.de>
Cc: "Ernst Christen (christen.1858@comcast.net)" <christen.1858@comcast.net>


Hi Joachim Hasse,

I am including the vhdl-ams reflector in the cc, since your point is really
interesting and I think that it may bring up more discussions about the
Q'DOT and other language issues.

It seems like that your interpretation about Q'DOT in the case of
discontinuities of Q *is zero at all times*.

I looked at the LRM again and Q'DOT seems to be the instantaneous value of
Q's derivative relative to time at any given time point.
So my interpretation for Q'DOT *is a Dirac function at the t0 and all zero
at other times*. My perspective is a pure mathematical point of view.

So the key question is that how should be deal with this kind of Dirac
Q'DOT?
a) Ignore it and treat it as zero:
        (My opinion: 1. Q'slew don't have the desired effect under this
interpretation.
                             2. Will it be a concern that Q'DOT value is
not defined in this case?)
b) treat it as a very very large value (Dirac function):
        (My opinion: It leads to my previous observation where current
Q'SLEW behavior in LRM is problematic)
c) Don't allow Q to be discontinuous
        (My opinion: too restrictive on quantity)
d) any other interpretation/suggestions .... ???

Since I am not familiar with the history of Q'DOT and Q'SLEW, there may be
some consideration or assumptions which I am not aware of.
So please bring up your questions or two cents even if you didn't fully
understand the issue. It will be a good learning experience and some fun
mind exercise for all of us.

Best Regards,
Zhichao



On Tue, Sep 11, 2012 at 3:29 AM, Joachim Haase <
Joachim.Haase@eas.iis.fraunhofer.de> wrote:

>  Hi Zhichao,****
>
> ** **
>
> I think the basic problem is how to determine Q’DOT in your example. To my
> opnion, at the discontinuity the left-hand side value and the right-hand
> side value are 0.0. Thus, it seems to me that in the case of your example
> *Q’DOT equals 0.0 for all times.* Then I expect, that Q2 should follow Q
> as shown in the attached PDF. It seems at a first glance that this behavior
> does not meet the intuition. However, it seems to be in accordance with the
> current LRM. There is a difference on applying ´SLEW on a quantity and a
> signal at a discontinuity. L
>
> ** **
>
> The crucial problem seems to be to determine the right hand-side value
> Q’DOT at the discontinuity if  Q’DOT is not part of the equations that
> describe the simulation problem. I miss in the LRM how to determine Q’DOT
> in this case if Q’SLEW is applied. ****
>
> ** **
>
> Kind regards,****
>
> Joachim  ****
>
> ** **
>
> *Von:* owner-vhdl-ams@eda.org [mailto:owner-vhdl-ams@eda.org] *Im Auftrag
> von *ZHICHAO DENG
> *Gesendet:* Montag, 10. September 2012 19:19
> *An:* vhdl-ams@eda.org
> *Betreff:* Q'slew behavior with a discontinuous Q****
>
> ** **
>
> Hi all,
>
> I have come across with an issue related to the correct behavior of Q'slew
> in VHDL-AMS standard.
>
> For the following pseudo example,
> quantity Q, Q2 real :=0.0;
> if (xxx)
>     Q == 1.0;
> else
>     Q == 2.0;
>
> Q2 == Q'slew(100, -100);
>
> Suppose, quantity Q has an immediate jump at t0 from 1.0 to 2.0;
>
> xxx: some digital condition.****
>
> Q, Q2 are quantities.****
>
> ** **
>
> The attached doc contains my interpretation of the LRM behavior with
> illustrated figure. I also explain what may be the should-be behavior in
> this case and another option to deal with this if my interpretation of LRM
> is accurate.****
>
> ** **
>
> Please share your thoughts on this issue.****
>
> ** **
>
> Regards,****
>
> Zhichao****
>
> ** **
>
>
> --
> This message has been scanned for viruses and
> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
> believed to be clean. ****
>

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.




To unsubscribe to the vhdl-ams mailing list:
mailto:Majordomo@eda.org?subject=Unsubscribe&body=unsubscribe%20vhdl-ams