[SI-LIST] Re: NEXT and FEXT: Question on relative levels

  • From: "John Ellis" <John.Ellis@xxxxxxxxxxxx>
  • To: <Chris.Cheng@xxxxxxxxxxxx>, <si-list@xxxxxxxxxxxxx>
  • Date: Wed, 30 Nov 2005 17:14:00 -0800

Hi Chris,

FEXT and NEXT are transmission line phenomena and are not particular to
any signaling technology.

I think basically what happens in a differential pair is you begin to
excite other modes.  I'm still noodling on this so if someone wants to
jump in and correct me please do so.

With a differential signal you have=20

     XXXXX          XXXXXX
       1              -1    I think this would break down into 4 modes
      =20
1      1/2             1/2
2      1/2            -1/2
3     -1/2            -1/2
4      1/2            -1/2

In a truly differential signal, I think the "even" modes, 1 and 3 should
cancel out leaving only the "odd" modes 2 and 4.  With a less than
balanced differential pair the 'even' modes don't completely cancel and
common mode noise is generated.

As far as the terminations apply to FEXT and NEXT, I think the purpose
would be to keep Near end noise from reflecting and showing up at the
far end and vice versa.

John



John Ellis
Sr. Staff R&D Engineer
=20
Synopsys, Inc.

=20

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
On Behalf Of Chris Cheng
Sent: Wednesday, November 30, 2005 7:32 PM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: NEXT and FEXT: Question on relative levels

For some reason I always relate FEXT and NEXT with GigE. How does these
=3D
even and odd mode work with differential signals ? Or with current mode
=3D
drivers with matched impedance terminators at driver and receiver ends ?

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx]On Behalf Of John Ellis
Sent: Wednesday, November 30, 2005 4:19 PM
To: howie03@xxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: NEXT and FEXT: Question on relative levels


Hi Howard,

Long ago I was taught to think of far end crosstalk as the difference in
modal velocities.  If you have two lines, 1 active, 1 quiet  1 and 0
below, you have an even and an odd mode.

Aggressor        Victim
  XXXXX          XXXXX

    1              0


    1/2           1/2   EVEN Mode

    1/2          -1/2   ODD Mode

I will ignore NEXT and assume it is not reflected at all.

As a signal propagates down the line, in a microstrip, the odd modes
travel faster, so on the victim line, the far end crosstalk signal
begins to go negative , since the odd mode arrives first.  It will
continue to go negative until the even mode component on the victim line
arrives, at which time the pulse will head in the positive direction
until it reaches 0.  If the line is long enough, the difference in
arrival time of the two modes will be so great that the noise will reach
-1/2 or 50% and then stay there until the even mode component arrives.

In a stripline, the even and odd modes travel through the same medium,
so the +1/2 and -1/2 components on the victim line arrive at the same
time, yielding 0 far enc crosstalk.=3D3D20

John Ellis
Sr. Staff R&D Engineer
=3D3D20
Synopsys, Inc.

=3D3D20

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
On Behalf Of Dr. Howard Johnson
Sent: Wednesday, November 30, 2005 6:30 PM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: NEXT and FEXT: Question on relative levels

Hi David,

I will comment on your question, if you can excuse my
previous confusion -- having one of those days, I guess,
something about the construction of this question just seems
to be getting to me..., and I replied to quickly to someone
who had written to me...  anyway we are discussing
cancellation of far-end-crosstalk FEXT and whether therefore
in a microstrip the near-end-crosstalk NEXT can exceed (did
I get it right this time?) the FEXT, and I said yes, it can
and does in many cases.

I do not have any information about the relative rms values,
but as far as I can tell the AREAS under the FEXT and NEXT
curves seem to hold a constant ratio that is determined by
the geometry of the transmission structure. This ratio does
NOT vary with line length or risetime. The AREA for both
effects grows in direct proportion to the length of the
line.

I should mention also that the FEXT amplitude does not grow
forever with increasing length. If you project a
transmission line all the way to Pluto, with FEXT growing
strictly in proportion to length, the predicted FEXT
amplitude would grow unreasonably large; what seems to
happen instead is that the FEXT progresses to a level of 50%
after which you see the duration of the FEXT effect begin to
extend, much as with NEXT.

Best regards,
Dr. Howard Johnson, Signal Consulting Inc.,
tel +1 509-997-0505,  howie03@xxxxxxxxxx
http:\\sigcon.com  -- High-Speed Digital Design seminars,
publications and films


-----Original Message-----
From: David Siadat (dsiadat) [mailto:dsiadat@xxxxxxxxx]
Sent: Wednesday, November 30, 2005 2:35 PM
To: howie03@xxxxxxxxxx
Subject: RE: [SI-LIST] Re: NEXT and FEXT: Question on
relative levels


Dr. Johnson,

Would you comment on the difference between peak versus rms
noise of the
example you have shown.
I was wondering if the area under each (NEXT and FEXT) plot
has any
merit in noise comparison?

Regards,
David

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx]
On Behalf Of Dr. Howard Johnson
Sent: Wednesday, November 30, 2005 2:21 PM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: NEXT and FEXT: Question on relative
levels

To Charles et. al.,

I will make the claim that NEXT can and does often exceed
FEXT.  In a
homogeneous dielectric (same Er all around), with a uniform
cross-section trace I believe you always get perfect
cancellation of
FEXT.  When you mix some air into the configuration (as in a
microstrip)
it weakens the capacitive coupling and you always get (as
far as I know)
a negative FEXT coefficient. This last statement is an
assertion based
on experience and measurement, but not something for which I
have a
proof. I would be grateful to anyone who can point me to a
solid proof
of that assertion.

In ordinary pcb applications, if you plot crosstalk
amplitude versus
line length, the NEXT and FEXT waveforms start out with NEXT
biggest,
and they keep the same ratio as they both grow in proportion
to line
length. Then, once the line becomes long enough so that the
roundtrip
delay exceeds the signal rise/fall time (this is a "critical
length" for
the NEXT/FEXT problem), NEXT stops growing in amplitude but
instead
extends in duration.  FEXT keeps growing in amplitude,
eventually
exceeding the NEXT amplitude.

I have an example pictures here:
www.sigcon.com/Pubs/misc/TemporaryPicture.gif
The picture was computed with Hyperlynx linesim. I show this
picture,
and some others like it, in my High-Speed Digital Design
seminar (next
one is in January 2006).

The picture shows both NEXT and FEXT waveforms (measured at
opposite
ends of a victim line with terminations at both ends), for
various
lengths of 1/2, 1, 2, 4, 8 and 16 inches.
The signal risetime is 1 ns.
The microstrip trace geometry appears in the picture.
Height=3D3D3D7 mils, Width=3D3D3D7 mils, Separation (edge to edge) =
=3D3D3D 15
mils, Z0=3D3D3D64
ohms, on an FR-4 type substrate.
The crossover point (where NEXT=3D3D3DFEXT) for this particular
setup occurs
at about 11 inches length. Crosstalk of either type has a
peak amplitude
of about 8% at that point.  Below this critical length, NEXT
is bigger.

A faster risetime would exaggerates the FEXT, making it
larger compared
to NEXT and moving the crossover point to a shorter
distance.  For
example, with a 100-ps rise/fall time the crossover would
happen at
about 1 inch, with crosstalk of both types at 8% at that
risetime and
length.

Comments?

Best regards,
Dr. Howard Johnson, Signal Consulting Inc., tel +1
509-997-0505,
howie03@xxxxxxxxxx http:\\sigcon.com  -- High-Speed Digital
Design
seminars, publications and films


-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx]On Behalf Of
Aubrey_Sparkman@xxxxxxxx
Sent: Wednesday, November 30, 2005 8:05 AM
To: chengyuming_ah@xxxxxxxxxxxx;
Charles.Grasso@xxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: ??: NEXT and FEXT: Question on
relative levels


Let's be a little more clear about the situation we are
discussing.  For
=3D3D3D stripline I would expect (due to cancellation) NEXT to be
greater than
=3D3D3D FEXT.

BUT!

Charles asked about two MICROSTRIP traces.  Anyone want to
claim the =3D3D3D
inductive and capacitive coupling for microstrip traces is
sufficiently
=3D3D3D balanced for cancellation to apply?

Aubrey Sparkman=3D3D3D20
Enterprise Engineering Signal Integrity Team Dell, Inc.=3D3D3D20
Aubrey_Sparkman@xxxxxxxx=3D3D3D20
(512) 723-3592

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx] =3D3D3D
On Behalf Of Yuming Cheng
Sent: Tuesday, November 29, 2005 11:29 PM
To: Charles.Grasso@xxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] ??: NEXT and FEXT: Question on relative
levels

Hello Charles,

>From figure 3.2 of "High-Speed Digital System Design=3D3D3DA1=3D3D3DAAA
Handbook of =3D3D3D
Interconnect Theory and Design Practices", you can find that
the =3D3D3D
inductive and capacitive crosstalk cancel at far end and
reinforce at =3D3D3D
near end.

You can also get this from section 5.7.1-5.7.3 in
"High-speed digital =3D3D3D
design, Dr. Johnson".
 =3D3D3D20
So it's possible to have NEXT greater than FEXT.

B.R.
Astrom

--- "Grasso, Charles"
<Charles.Grasso@xxxxxxxxxxxx>=3D3D3DD0=3D3D3DB4=3D3D3DB5=3D3D3DC0:

> Is it possible to have NEXT greater than FEXT.??
> =3D3D3D20
>=3D3D3D20
> For example: For a two trace microstrip system my
simulator reports=3D3D3D20
> 10% NEXT (so for a 2.5V signal that translates to 250mV).
>=3D3D3D20
> Based on the reported mutuals and self numbers and using
the formula=3D3D3D20
> in Eric's book I estimate 60mV for FEXT.
>=3D3D3D20
> =3D3D3D20
>=3D3D3D20
> Is this reasonable?
>=3D3D3D20
> =3D3D3D20
>=3D3D3D20
> Best Regards
> Charles Grasso
> Senior Compliance Engineer
> Echostar Communications Corp.
> Tel: 303-706-5467
> Fax: 303-799-6222
> Cell: 303-204-2974
> Pager/Short Message: 3032042974@xxxxxxxx
> Email: charles.grasso@xxxxxxxxxxxx;
> Email Alternate: chasgrasso@xxxxxxxx
>=3D3D3D20
> =3D3D3D20
>=3D3D3D20
>=3D3D3D20
>
------------------------------------------------------------
------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the
Subject field
>=3D3D3D20
> or to administer your membership from a web page, go
> to:
> //www.freelists.org/webpage/si-list
>=3D3D3D20
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject
field
>=3D3D3D20
> List FAQ wiki page is located at:
>               =3D3D3D20
> http://si-list.org/wiki/wiki.pl?Si-List_FAQ
>=3D3D3D20
> List technical documents are available at:
>                 http://www.si-list.org =3D3D3D20
> List archives are viewable at:    =3D3D3D20
>               //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
>  =3D3D3D20
>=3D3D3D20
>=3D3D3D20



=3D3D3D09

=3D3D3D09
        =3D3D3D09
___________________________________________________________
=3D3D3DD1=3D3D3DC5=3D3D3DBB=3D3D3DA2=3D3D3DC3=3D3D3DE2=3D3D3DB7=3D3D3DD1G=
=3D3D3DD3=3D3D3DCA=3D
3D=3D
3DCF=3D3D3DE4=3D3D3DA3=3D3D3DAD=3D3D3D=3D3D
D6=3D3D3DD0=3D3D3DB9=3D3D3DFA=3D3D3DB5=3D3D3DD
A=3D3D3DD2=3D3D3DBB=3D3D3DBE=3D3D3DF8=3D3D3D
=3D3D3DCE=3D3D3DDE=3D3D3DC0=3D3D3DAC=3D3D3DBB=3D3D3DF8=3D3D3DD3=3D3D3DCA=3D=
3D3DBC=3D3D3DFE=3D3
D=3D
3DC9=3D3D3DA7=3D3D3DC8=3D3D3DC5=3D3D3D=3D3D
B3=3D3D3DAC=3D3D3DB4=3D3D3DF3=3D3D3DD3=3D3D3DCA
=3D3D3DCF=3D3D3DE4
http://cn.mail.yahoo.com

------------------------------------------------------------
------
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:    =3D3D3D20
                //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
 =3D3D3D20
------------------------------------------------------------
------
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:    =3D3D20
                //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
 =3D3D20

------------------------------------------------------------------
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:    =3D20
                //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
 =3D20

------------------------------------------------------------------
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
  

Other related posts: