[SI-LIST] Re: (no subject)

My 2(euro)cents.
I understand that dispersion in a digital/analogue signal transmission path
means that there are more modes available in the line which will account for 
change in delay and/or phase.
The example of the optical fibre gives
a)      Multimode: the launch angle of the source is random and each angle 
will lead to different attenuation and delay. And thus dispersion.
b)      Single mode fibre: this has a reduced aperture so that there are very 
limited accepted angles of launch and therefore minimal dispersion.
c)      Graded index:   This has a wider aperture allowing more angles of launch
but the relative refraction is varied to help compensate the various angles and 
thus reduce the dispersion.

In a multilayer board it is possible to have "odd" and "even" modes especially 
with balanced configurations that may well have difference in losses and 
medium and thus dispersion. Generally there is a 10:1 ratio in power.
Other modes such as "parallel plate" are generally negligible.
In FR4 it is possible to have frequency dependance in loss and Er.

As pointed out by previous members the skin-effect and for higher frequencies
the "roughness" are frequency dependent phase and amplitude affecting.
-- 
Regards,    __________   James G Roberts
           /___  ____ |  jrobert@xxxxxxxxxxxxxxxxxxxxx
 Jim      __  / /___/ /  jgroberts@xxxxxxxxxx
         / /_/ /---| |   Room: BE436, Hilversum
         \____/    /_/   Tel: +31 35 687 4308 Fax: 5976

======================================================================================
Mike wrote:
> 
> I would try to think of it as "dispersing" the "intelligence" of the signal...
> intelligence = both magnitude and phase
> 
> Todd Westerhoff wrote:
> 
> > I've always understood dispersion as the effect of different velocities at
> > different frequencies, effectively scattering or "dispersing" the arrival
> > time of the signal wavefront.
> >
> > Todd.
> >
> > Todd Westerhoff
> > High Speed Design Specialist
> > Cisco Systems
> > 1414 Massachusetts Ave - Boxboro, MA - 01719
> > email:twesterh@xxxxxxxxx
> > ph: 978-936-2149
> > ============================================
> >
> > "When did the choices get so hard, with so much more at stake?
> >  Life gets mighty precious when there's less of it to waste"
> >
> > - Bonnie Raitt, "Nick of Time"
> >
> > -----Original Message-----
> > From: si-list-bounce@xxxxxxxxxxxxx
> > [mailto:si-list-bounce@xxxxxxxxxxxxx]On Behalf Of Yu Liu
> > Sent: Monday, December 09, 2002 6:19 PM
> > To: Loyer, Jeff
> > Cc: si-list@xxxxxxxxxxxxx
> > Subject: [SI-LIST] Re: (no subject)
> >
> > Jeff,
> >
> > By definition, I think dispersive effect is the frequency-related effect
> >   (phase velocities at different spectrums). However, since the losses
> > are also frequency-dependent, sometimes people just use the term
> > interchangbly, which causes confusions. Distinguish the two terms
> > (dispersive and loss) would make things clearer.
> >
> > Yu
> > ===
> >
> > Loyer, Jeff wrote:
> >
> > > Yu,
> > > When you use the term "dispersive", are you talking about losses
> > (resistive, skin effect, dielectric), or about differences in phase
> > velocities (page 170 of Pozar's book)?  I've heard others refer to loss
> > effects as dispersive and have had confusion as a result.  Are both uses of
> > the term "dispersive" correct?
> > >
> > > Jeff Loyer
> > >
> > >
> > > -----Original Message-----
> > > From: Yu Liu [mailto:yu_liu@xxxxxxxxxxxxx]
> > > Sent: Friday, December 06, 2002 9:32 PM
> > > To: xlzhou@xxxxxxxxx
> > > Cc: 'Xin Wu'; si-list@xxxxxxxxxxxxx
> > > Subject: [SI-LIST] Re: (no subject)
> > >
> > >
> > >
> > > Mick,
> > >
> > > To answer your question: In theory, Yes. In practice, No.
> > >
> > > The fundamental principle for SPICE model approach is using lumped
> > > elements (RLC) to model distributed effect (S-parameter). For lower
> > > frequency and small size, it is possible. But for broad-band frequency
> > > and large size (say, 20' FR-4), it is very hard to model the dispersive
> > > effects accurately. That's why people start to use frequency-dependent
> > > elements, or better yet, import S-parameter directly into Spice engine.
> > >
> > >
> > > Regards,
> > >
> > > Yu
> > > ===
> > >
> > > Apache Design Solutions
> > > web: www.apache-da.com
> > >
> > >
> > >
> > > Zhou, Xingling (Mick) wrote:
> > >
> > >
> > >>Xin Wu
> > >>
> > >>Thanks for your information. It will be valuable.
> > >>However, it still contains something like (from the example on the web you
> > >>referred)
> > >>
> > >>E11 11 12 FREQ {V(10,5)}=
> > >>
> > >>which will not be accepted by some circuit simulators such as ADS. It is
> > >>similar to Ansoft fullwave SPICE. I know HSPICE and PSPICE can handle it.
> > >>
> > >>Although it is still a research topic in terms of efficiency, stability
> > and
> > >>accuracy etc., I know the problem [S] ---> SPICE model (reduced models) is
> > >>solvable without the following limitations in several ways. But, can we
> > >>generate broadband SPICE models without frequency dependent lookup tables
> > >>such as " FREQ " ?  Or is it still solvable without components "FREQ" etc.
> > ?
> > >>
> > >>Regards,
> > >>
> > >>Mick
> > >>
> > >>-----Original Message-----
> > >>From: Xin Wu [mailto:lifehappiest@xxxxxxxxxxx]
> > >>Sent: Friday, December 06, 2002 8:38 PM
> > >>To: scott@xxxxxxxxxxxxx; xlzhou@xxxxxxxxx
> > >>Cc: si-list@xxxxxxxxxxxxx
> > >>Subject: Re: [SI-LIST] Re: (no subject)
> > >>
> > >>
> > >>Hi, XL,
> > >>   If you want to convert S-matrix into a spice simulator, a lot of
> > >>commercial software can do that--that's the simplest way. However, it is
> > >>possible for yourself to make a equivalent subcircuit which can include
> > the
> > >>timing and frequency info and then import to a spice like simulator. Here
> > >>are some application notes:
> > >>http://wwwinfo.cern.ch/ce/ae/Maxwell/apps/2stripem/2stripem.html
> > >>If you want to get involved in more details, I remembered that somebody
> > did
> > >>his Ph.D work on this in UIUC around early 1990s'.
> > >>
> > >>
> > >>
> > >>Xin Wu
> > >>Ph.D student,
> > >>University of Maryland, College Park
> > >>Http://www.wam.umd.edu/~xwu
> > >>
> > >>
> > >>
> > >>
> > >>
> > >>
> > >>
> > >>>From: "Scott McMorrow" <scott@xxxxxxxxxxxxx>
> > >>>Reply-To: scott@xxxxxxxxxxxxx
> > >>>To: xlzhou@xxxxxxxxx
> > >>>CC: "'si-list@xxxxxxxxxxxxx'" <si-list@xxxxxxxxxxxxx>
> > >>>Subject: [SI-LIST] Re: (no subject)
> > >>>Date: Fri, 06 Dec 2002 13:45:41 -0800
> > >>>
> > >>>
> > >>>Mick,
> > >>>
> > >>>There are two solutions which I use:
> > >>>
> > >>>1) Apache Nspice is an Hspice compatabible simulator that can
> > >>>co-simulate with Touchstone S-parameter files.
> > >>>http://www.apache-da.com/
> > >>>
> > >>>2) Sigrity BroadBand Spice can convert Touchstone s-parameter files into
> > >>>spice black boxes.
> > >>>http://www.sigrity.com/
> > >>>
> > >>>Both are scary accurate in both the frequency and time domain.
> > >>>
> > >>>regards,
> > >>>
> > >>>scott
> > >>>
> > >>>--
> > >>>Scott McMorrow
> > >>>Teraspeed Consulting Group LLC
> > >>>2926 SE Yamhill St.
> > >>>Portland, OR 97214
> > >>>(503) 239-5536
> > >>>http://www.teraspeed.com
> > >>>
> > >>>
> > >>>Zhou, Xingling (Mick) wrote:
> > >>>
> > >>>
> > >>>
> > >>>>Hello,
> > >>>>
> > >>>>Did anybody successfully generate broadband (DC-20GHz) SPICE model from
> > S
> > >>>>matrix for complex structures using the SPICE generator in ADS ? Or any
> > >>>>other alternatives. I know Ansoft fullwave SPICE does the work.
> > >>>>Unfortunately, some circuit simulators do not support the frequency
> > >>>>dependent lookup table sources. We can definitely use [S] directly, but
> > >>>>
> > >>>>
> > >>>some
> > >>>
> > >>>
> > >>>>users still prefer SPICE models. Any experience to share ?
> > >>>>
> > >>>>Thanks,
> > >>>>
> > >>>>Xingling(Mick) Zhou, PhD
> > >>>>Signal Integrity Technologist
> > >>>>Agere Systems
> > >>>>
> > >>>>Tel: 610-712-7462
> > >>>>Fax: 610-712-4081
> > >>>>
------------------------------------------------------------------
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 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: