Subject: RE: net_resolution
From: Jonathan Sanders (jons@cadence.com)
Date: Mon Feb 26 2001 - 22:05:13 PST
Ian,
Thanks for the reply. I think we are now in agreement on where this should
occur the only remaining question is on the forcing of the bidir to support
this.
My question is one that I pondered with Bill Hobson several times regarding
the "view" that driver-receiver segregation means bidir connect
modules. My initial
thought is that bidir should be used in cases where the design has a bidir
port versus
where a net has both drivers and receivers.
I say this as a bidir port in a design is not writing and reading at the
same time (at
least most of the ones I have seen and of course analog) but switching back and
forth and thus a special connect module is needed. In this case we are not
really
doing a bidir conversion but what I call a duel direction conversion, the
digital port
is duel directional.
If we require that bidir connection modules be used in all driver-receiver
segregation
cases then we essentially get to what Bill and Don wrote in their appnote
provided
to the committee many months ago: all connect modules end up being bidir,
ok not
all but most. The number of mixed nets that have both drivers and
receivers I suspect
will be quite high and these would then need bidir.
I suspect you propose it to be bidir since it reads and writes to the
digital port
of the connect module, is this your reason? I suspect we will need to have
further
discussions on this in our face to face meeting but hopefully you see my issues
that I am wrestling with.
Jon
At 02:53 PM 2/26/01, Ian Wilson wrote:
> > -----Original Message-----
> > From: Jonathan Sanders [mailto:jons@cadence.com]
> > Sent: Monday, February 26, 2001 8:19 AM
> >
> >
> > It seems that you are implying that the only time driver receiver
> segregation
> > occurs is if the user provides a rule that uses a bidir connect
> module. Thus
> > the only cases where this can happen are:
> >
> > 1) A true digital inout port exists on the net
> > 2) Coercion is used and the net is declared digital, thus the CM
> goes across
> > the analog port that is inout
>
>My statement was incorrect. Whenever there are both digital drivers and
>digital
>loads on a mixed net, the inserted connect module should be bidrectional (i.e.
>have a digital inout port).
>
> >
> >
> > The additional clarification is that you are now forcing the CM to
> use a
> > inout port
> > rather than allowing any CM to support driver-receiver seg.
> >
> > Is there a reason why analog effects on the digital receivers
> should be
> > limited
> > to such conditions? Should not the digital drivers always be
> influenced by
> > analog before the digital drivers are driven?
>
>Agree.
>
>Side note: obviously there need to be both drivers and receivers on a net
>in order
>for there to be anything to segregate.
>
>--ian
***********************************************************
Jonathan L. Sanders
Product Engineering Director
Mixed Signal and Physical Verification Solutions
Cadence Design Systems, Inc.
555 River Oaks Pkwy
San Jose, CA. 95134
INTERNET:jons@cadence.com Tel: (408) 428-5654 Fax : (408) 944-7265
***********************************************************
This archive was generated by hypermail 2b28 : Mon Feb 26 2001 - 22:08:37 PST