[SI-LIST] Re: Question regarding return current in a differential pair
- From: steve weir <weirsi@xxxxxxxxxx>
- To: glenchen@xxxxxxxxxxxxx,<si-list@xxxxxxxxxxxxx>
- Date: Fri, 21 Apr 2006 02:54:07 -0700
Glen, no that is not correct. When you tightly couple, you need to
manage both the even and odd mode impedances throughout the
path. Loosely coupled, the energy coupled between the lines is so
insignificant that it can be virtually ignored. That
simplifies: routing worries, considerations for transitions through
connectors, and termination complexity.
Steve.
At 11:21 PM 4/20/2006, glenchen@xxxxxxxxxxxxx wrote:
>Hi, Steve
> I am novice of SI principle. But i have different view
> from your word "
>All things considered, if you have the space, loose
>coupling is easier to make work right. " . If the differential traces
>are laied tightly, the environment they see is more like the same (
>whatever dielectric
>property , dielectric thickness...) , so i think "tight coupling is
>easier to make things work .
>Am i right ?
>
>Regards
>- Glen Chen / SynerChip Tech.
>
>
>steve weir <weirsi@xxxxxxxxxx>
>Sent by: si-list-bounce@xxxxxxxxxxxxx
>
>04/21/2006 12:38 PM
>Please respond to
>weirsi@xxxxxxxxxx
>
>To
>bfields@xxxxxxxx, <Scott.Nixon@xxxxxxxxxxxx>, <si-list@xxxxxxxxxxxxx>
>cc
>Subject
>[SI-LIST] Re: Question regarding return current in a differential pair
>
>
>
>
>Brian, I agree on the value of Eric's book. But I think you are
>drawing very different conclusions from it than I do. Consider that
>if 20% ( n optimistic number ) of the energy is coupled between the
>two lines then that is also the extent of cross talk attenuation the
>pair exhibit against an aggressor versus just one or the other
>line. I think that debunking the notion that a diff pair on a PCB
>markedly improves either radiation or susceptibility is one of Lee's
>favorite topics for good reason.
>
>A diff pair that doesn't skew too badly carries its own switching
>reference, is DC neutral and has twice the signal energy for a given
>constant p-p voltage swing on each of the constituent signals. Those
>are each compelling features by themselves. Together, they are hard
>to beat. Opinions vary on the value of tight versus loose
>coupling. All things considered, if you have the space, loose
>coupling is easier to make work right. You still want to keep each
>trace in the pair in the same electrical environment. It has been my
>experience that tight coupling is primarily born from density
>requirements, not performance superiority or ease of implementation.
>
>Steve
>
>At 05:18 PM 4/20/2006, Fields, Brian wrote:
> >Scott,
> >
> >Tightly coupled pairs (if we can consider 10-20% coupling between traces
> >as tight) have other advantages besides providing a current return path.
> >
> >
> >Keeping the traces close together helps cancel out crosstalk induced by
> >a nearby aggressor net. If the pair is spaced 5 mil edge-to-edge, then
> >it's reasonable to assume the crosstalk magnitude will be approximately
> >equal on each line. If the pair is spaced far apart then the crosstalk
> >on one line will be greater than on the other, as the magnitude falls
> >off with distance, and the noise is no longer common mode; now there is
> >noise on your differential signal.
> >
> >Another benefit of keeping differential pairs close is that they're just
> >easier to route, making it easier to keep the trace lengths equal and
> >reduce skew.
> >
> >(Eric Bogatin's book has an extensive chapter on differential pairs that
> >I've used as a reference in the past, it addresses the coupling issue in
> >some detail.)
> >
> >- B
> >
> >-----Original Message-----
> >From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
> >On Behalf Of Scott.Nixon@xxxxxxxxxxxx
> >Sent: Thursday, April 20, 2006 5:29 PM
> >To: si-list@xxxxxxxxxxxxx
> >Subject: [SI-LIST] Re: Question regarding return current in a
> >differential pair
> >
> >Tom,
> >
> >If the coupling between a differential pair is only about 20% on a PCB,
> >are the
> >benefits of using differential pairs really that significant? The signal
> >is
> >still 80% coupled to the reference plane. I'm a student and I just want
> >an idea
> >of how this is justified in a practical application.
> >
> >Much Thanks,
> >Scott Nixon
> >------------------------------------------------------------------
> >To unsubscribe from si-list:
> >si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
> >
> >or to administer your membership from a web page, go to:
> >//www.freelists.org/webpage/si-list
> >
> >For help:
> >si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
> >
> >List FAQ wiki page is located at:
> > http://si-list.org/wiki/wiki.pl?Si-List_FAQ
> >
> >List technical documents are available at:
> > http://www.si-list.org
> >
> >List archives are viewable at: =20
> > //www.freelists.org/archives/si-list
> >or at our remote archives:
> > http://groups.yahoo.com/group/si-list/messages
> >Old (prior to June 6, 2001) list archives are viewable at:
> > http://www.qsl.net/wb6tpu
> > =20
> >
> >------------------------------------------------------------------
> >To unsubscribe from si-list:
> >si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
> >
> >or to administer your membership from a web page, go to:
> >//www.freelists.org/webpage/si-list
> >
> >For help:
> >si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
> >
> >List FAQ wiki page is located at:
> > http://si-list.org/wiki/wiki.pl?Si-List_FAQ
> >
> >List technical documents are available at:
> > http://www.si-list.org
> >
> >List archives are viewable at:
> > //www.freelists.org/archives/si-list
> >or at our remote archives:
> > http://groups.yahoo.com/group/si-list/messages
> >Old (prior to June 6, 2001) list archives are viewable at:
> > http://www.qsl.net/wb6tpu
> >
>
>------------------------------------------------------------------
>To unsubscribe from si-list:
>si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
>or to administer your membership from a web page, go to:
>//www.freelists.org/webpage/si-list
>
>For help:
>si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
>List FAQ wiki page is located at:
> http://si-list.org/wiki/wiki.pl?Si-List_FAQ
>
>List technical documents are available at:
> http://www.si-list.org
>
>List archives are viewable at:
> //www.freelists.org/archives/si-list
>or at our remote archives:
>
>http://groups.yahoo.com/group/si-list/messages
>Old (prior to June 6, 2001) list archives are viewable at:
> http://www.qsl.net/wb6tpu
>
>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
