[SI-LIST] Re: Question on varying the coupling ratio on a differential pair

Hi Bill -- I'm not totally comfortable with your characterization.  We
both come to the same conclusion, so we might be trying to say the same
thing.

I would say that in the regime you describe -- 100 ohms differential
impedance from point A to point B, no common mode component in the
signal -- there will be no reflections by definition.  This follows
directly from TEM transmission line theory.  This can be proven from a
text book -- no measurements necessary.

The reason that there will be reflections in real life is that a system
such as that described in the original post -- going through connectors
from backplanes to cards, from tightly coupled to loosely coupled, etc.,
is not a TEM transmission line.  A signal path switching reference
planes, or for that matter exhibiting any non-uniformity in the
direction of propagation, violates the assumptions of TEM transmission
line theory.  Without transmission line theory the concept of a signal
path even having a characteristic impedance starts to get flimsy and
must be used with care.  So the issue isn't that we have a constant
impedance 100-ohm transmission line exhibiting reflections -- it is that
we don't even have a transmission line in the rigorous sense.

The original post is an example of a conclusion drawn by extrapolating
transmission line theory beyond its range of validity.  If we treat a
connector as a 100 ohm differential pair connecting two other 100-ohm
differential pairs, we predict no reflections when the system is excited
differentially.  Once we realize that the connector (despite the best
efforts of its designers) is not a TEM transmission line, it becomes
clear that we should expect reflections in the system described in the
original post for any reasonably current data rate.  As you point out,
whether they actually matter will depend entirely on the application.

I should add, for clarity, that this is not an indictment of TEM (or
quasi-TEM) transmission line analysis as a tool for circuit analysis and
design, even at today's speeds.  However, just like any modeling
technique, using its results outside of its range of validity will
inevitably lead to grief.

  -- Steve


------------------------------------------------------------------------
-
Steven D. Corey, Ph.D.
Principal Engineer
Tektronix - Enabling Innovation
=20
http://www.tdasystems.com
http://www.tektronix.com
=20
email: steven.corey@xxxxxxxxxxxxx
phone: (503) 627-6816
fax:   (503) 627-2260
------------------------------------------------------------------------
-
=20

>-----Original Message-----
>From: si-list-bounce@xxxxxxxxxxxxx=20
>[mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Bill Wurst
>Sent: Monday, January 15, 2007 5:23 PM
>To: si-list@xxxxxxxxxxxxx
>Subject: [SI-LIST] Re: Question on varying the coupling ratio=20
>on a differential pair
>
>Wayne,
>
>Even if the differential impedance is 100 Ohms from the source to the=20
>final load and the signal is purely differential (no common-mode=20
>component, originating from either the source or skew within=20
>the pair),=20
>there will be reflections.  The question is, how significant are they.
>
>To better understand what is happening, you might refer to Dr.=20
>Johnson's=20
>second book, High-Speed Signal Propagation - Advanced Black Magic, pp.=20
>417-8, where he describes "the differential U-turn."  The current loop=20
>created in the "U-turn zone" has a magnetic field associated with it=20
>which can cause EMI, coupling into nearby conductors, and create an=20
>inductive discontinuity.
>
>If your transition times are slow enough relative to the=20
>electrical size=20
>of the "U-turn zone", you won't be able to "see" a reflection.
>
>Regards,
>     -Bill
>
>
>       /************************************
>      /         billw@xxxxxxxxxxx         /
>     /                                   /
>    / Advanced Electronic Concepts, LLC /
>   /           www.aec-lab.com         /
>   ************************************
>=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
>cookeway wrote:
>> Hi all,
>>=20
>> I have a question concerning differential pair coupling ratios that=20
>> is causing some debate here in our design group.  If a differential=20
>> pair with an impedance of 100 ohms connects through a backplane=20
>> connector and into a backplane such that the impedance is=20
>> consistently 100 ohms, but the coupling ratio of the return currents=20
>> on the adjacent planes versus the return current using the=20
>compliment=20
>> differential signal changes, will there be a reflection on=20
>the signal?
>>=20
>> For example, if you have a strongly coupled differential pair of 100=20
>> ohms on your motherboard which goes through a connector to=20
>connect to=20
>> a loosely coupled 100 ohm differential pair on the backplane, is=20
>> there a reflection?  My gut feeling is that the reflections=20
>caused by=20
>> the different coupling ratios would cancel each other out causing no=20
>> net reflection.  Am I right in thinking this?
>>=20
>> Thanks,
>>=20
>> Wayne Cooke
>> Hardware Engineer, Zeugma Systems Inc.
>> Suite 250, 13571 Commerce Parkway, Richmond, BC, Canada, V6V 2R2=20
>> Tel: (604) 247-2828=20
>> Fax: (604) 247-3251
>> wcooke@xxxxxxxxxxxxxxxxx=20
>>=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:
>http://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
>               http://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:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


List technical documents are available at:
                http://www.si-list.net

List archives are viewable at:     
                http://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: