RE: [SI-LIST] : Tools for high-speed interconnect design

About this list Date view Thread view Subject view Author view

From: Jian X. Zheng ([email protected])
Date: Mon Aug 14 2000 - 14:16:53 PDT


Hi, Everybody:

A suitable full-wave electromagnetic simulator can do much for PCB.
Certainly, for FEM and FDTD, the PCB structures can be simulated are still
much limited by the size. However, MOM electromagnetic simulators are
special for PCB structures. They can achieve much for PCB structures. I have
tried our IE3D MOM full wave electromagnetic simulator on multi-layer PCB
with 12 by 12 inches on my notebook computer. It still works quite well up
to 500 MHz. Interested si-engineers can contact me on this issue. Thanks!

Best regards,

-----------------------------------------------------------------------
Jian-X. Zheng, Ph.D
Zeland Software, Inc., 48890 Milmont Drive, 105D, Fremont, CA 94538, U.S.A.
Tel: 510-623-7162, Fax: 510-623-7135, Web: http://www.zeland.com
---------------------------------------------------------------------

> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of Fred Balistreri
> Sent: Monday, August 14, 2000 1:34 PM
> To: linda walling
> Cc: [email protected]
> Subject: Re: [SI-LIST] : Tools for high-speed interconnect design
>
>
> linda walling wrote:
> >
> > RE: [SI-LIST] : Tools for high-speed interconnect designDear Hassan,
> >
> > I agree with you that finite element simulation is not
> > practical by itself for simulating large circuits.
> > I don't believe that is the intent of HFSS V8. I
> > believe the correct usage of HFSS V8 would be to extract
> > parts of circuits that need full-wave simulation, then export
> > those models into SPICE to incorporate into larger models
> > that contain the rest of the circuit.
> >
> > What is meant by Full-Wave Spice is simply that you can take
> > structures (or parts of structures) that require full-wave 3D
> > simulation, and export a spice model. These structures can
> > be quite complex with coupled conductor terminals. They can
> > have lumped RLC elements contained inside the FEM model. But,
> > the engineer is still needed to intelligently create the full-circuit
> > spice model which incorporates the 3D FEM models along with all the
> > other typical spice models one might need to represent the full circuit.
> >
> > linda
> >
> > ********************************************************************
> > Linda Walling
> > High Frequency Applications Engineer
> > Southwest Region
> > ********************************************************************
> > Ansoft's U.S. customer support group for high-frequency products:
> >
> > Michael Brenneman (CA) 408-261-9095 [email protected]
> > Markus Kopp (FL) 954-571-0243 [email protected]
> > Liza Ma (CA) 310-375-7126 [email protected]
> > David Perry (NH) 603-643-5340 [email protected]
> > Richard Remski (TX) 817-516-7133 [email protected]
> > Steve Rousselle (CA) 310-937-1684 [email protected]
> > James Sherman (FL) 727-394-1023 [email protected]
> > John Silvestro (PA) 412-261-3200x127 [email protected]
> > Linda Walling (NM) 505-828-2103 [email protected]
> >
> > ***************************************************************
> > Ansoft's customer support for installation issues:
> > Louis Reash (PA), 412-261-3200x165 [email protected]
> >
> > ********************************************************************
> > Check out Ansoft's WWW site at http://www.ansoft.com for product
> > information, technical support, employment opportunities, etc.
> > ********************************************************************
> >
> > Haved you donated food today to a hungry person?
> > http://www.thehungersite.com/
> >
> > -----Original Message-----
> > From: Hassan Ali [mailto:[email protected]]
> > Sent: Monday, August 14, 2000 10:33 AM
> > To: linda walling; Charles Grasso; Ray Anderson; si-list
> > Subject: RE: [SI-LIST] : Tools for high-speed interconnect design
> >
> > Linda,
> >
> > Thanks for your response.
> >
> > For the benefit of the SI-list users it would be very appreciated if
> > you
> > could elaborate further on what you (Ansoft) mean by a "full-wave spice
> > tool". I personally take it to be a tool that is capable of performing
> > electronic circuit simulation using a full-wave approach i.e. a method
> > which
> > makes no assumptions about the nature and propagation of electromagnetic
> > field in the problem of interest. For example, it doesn't assume that
> > the
> > field is quasi-TEM (mostly in transverse direction and negligible axial
> > component). And that is exactly where I have a problem comprehending.
> >
> > In my opinion, to create a meaningful "full-wave spice" is not
> > practical.
> > Take for example a simple electronic circuit comprising of a voltage
> > source
> > and a one-inch of a microstrip line terminated with a 50-ohm resistor to
> > ground. To perform a full-wave simulation of that circuit (say with
> > Ansoft's
> > favourite Finite Element Method) you need to capture the material and
> > exact
> > geometry of the microstrip line and the resistor, and represent the
> > voltage
> > source vectorially (its magnitude and polarisation). Using FEM (or any
> > other
> > full-wave electromagnetic field solution technique), you will have to
> > discretize the geometry into smaller elements as the method dictates and
> > do
> > a time-domain solution. This solution process is practically impossible
> > with
> > thousands of circuits in a typical printed circuit board with thousands
> > of
> > components of various sizes and material compositions. The problem is
> > exacerbated by the presence of active components with non-linear
> > behaviour.
> > (Most electromagnetic field solution tools assume linearity in many
> > aspects). That's why I think it is virtually impossible (at least in the
> > next while) to have a true full-wave electronic circuit simulation tool.
> >
> > The idea that was floating around in academic circles about 5 years
> > ago
> > (when I last checked), was to find a way of formulating the problem in
> > such
> > away that the sections of the problem that lend itself well to
> > SPICE-like
> > solution (e.g. parts that have non-linear I/O devices) are formulated
> > using
> > SPICE-like techniques, and those sections that need 3D electromagnetic
> > field
> > solution (e.g. transmission lines, 3D vias, 3D packages) are handled
> > with 3D
> > electromagnetic field solution techniques. A mechanism were to be
> > created
> > for the two formulations to implicitly (internally) exchange inputs and
> > outputs and give rise to a single, unified matrix equation to be solved.
> > But
> > even that approach is far from being called a full-wave solution!
> >
> > I'd really appreciate knowing more about this matter especially new
> > developments in the last five years.
> >
> > Thanks.
> >
> > Hassan.
> >
> > -----Original Message-----
> > From: linda walling [mailto:[email protected]]
> > Sent: Friday, August 11, 2000 3:52 PM
> > To: Ali, Hassan [KAN:0G15:EXCH]; Charles Grasso; Ray Anderson;
> > si-list
> > Subject: RE: [SI-LIST] : Tools for high-speed interconnect design
> >
> > Hassan, Charles, & All,
> >
> > HFSS V8 will have a number of enhancements that make it a full-wave
> > spice tool.
> > The enhancements that will most directly focus on SI applications
> > include Modal-to-modal transformations, which enable a structure to
> > have
> > multiple
> > coupled conductors at the ports (or terminals), Spice model export
> > (to
> > Maxwell Spice,
> > PSPICE and HSPICE), lumped RLC elements,
> > and interpolative discrete sweeps, which will allow extremely large
> > problems to run
> > faster and smaller and allow for very accurate emissions testing.
> >
> > I am not sure when V8 will be released - I believe the scheduled
> > date is
> > around the end of
> > the year, but it has not been announced yet.
> >
> > linda
> >
> > ********************************************************************
> > Linda Walling
> > High Frequency Applications Engineer
> > Southwest Region
> > ********************************************************************
> > Ansoft's U.S. customer support group for high-frequency products:
> >
> > Michael Brenneman (CA) 408-261-9095 [email protected]
> > Markus Kopp (FL) 954-571-0243 [email protected]
> > Liza Ma (CA) 310-375-7126 [email protected]
> > David Perry (NH) 603-643-5340 [email protected]
> > Richard Remski (TX) 817-516-7133 [email protected]
> > Steve Rousselle (CA) 310-937-1684 [email protected]
> > James Sherman (FL) 727-394-1023 [email protected]
> > John Silvestro (PA) 412-261-3200x127 [email protected]
> > Linda Walling (NM) 505-828-2103 [email protected]
> >
> > ***************************************************************
> > Ansoft's customer support for installation issues:
> > Louis Reash (PA), 412-261-3200x165 [email protected]
> >
> > ********************************************************************
> > Check out Ansoft's WWW site at http://www.ansoft.com for product
> > information, technical support, employment opportunities, etc.
> > ********************************************************************
> >
> > Haved you donated food today to a hungry person?
> > http://www.thehungersite.com/
> >
> > -----Original Message-----
> > From: Hassan Ali [mailto:[email protected]]
> > Sent: Friday, August 11, 2000 1:33 PM
> > To: Charles Grasso; Ray Anderson; si-list
> > Cc: lwalling
> > Subject: RE: [SI-LIST] : Tools for high-speed interconnect design
> >
> > Charles,
> >
> > "Modes to nodes" thing won't give you a SPICE model from the way I
> > understood that concept from the Ansoft people and from my own past
> > research. I understood it to be a technique of coupling (interfacing)
> > the
> > traditional SPICE-based electrical circuit (nodal) simulation with 3D EM
> > field (modal) simulation. In this technique the models of the
> > interconnects
> > are implicitly generated and utilised. The generated interconnect models
> > would mean nothing to external SPICE-based simulators. Ansoft people may
> > correct me if I'm wrong.
> >
> > And BTW, is Ansoft HFSS 8 already available?
> >
> > Regards.
> I agree that a full wave solution can be used to simulate only certain
> parts to a PCB board. The solution is useful for investigating very
> high frequency effects or where non-TEM modes, dispersion and other
> losses can be accounted. However full wave SPICE to me is something
> different than described below. Full wave SPICE can be achieved using
> FDTD methods. Since the FDTD solution is done in the time domain
> integration
> with SPICE becomes much more feasible without loss of accuracy. Using
> frequency domain solvers envolves extracting out LCR or other circuit
> topologies from the S parameters. Getting a good fit for a broadband
> solution, as in digital circuits is always a challenge and may not be
> general in nature. To be sure there are challenges and problems with
> both approaches, however the FDTD solution seems a much more natural
> one. This is the approach we have taken.
>
>
> Best Regards,
> >
> > Hassan.
> >
> > -----Original Message-----
> > From: Charles Grasso [mailto:[email protected]]
> > Sent: Friday, August 11, 2000 2:28 PM
> > To: Ray Anderson; si-list
> > Cc: lwalling
> > Subject: Re: [SI-LIST] : Tools for high-speed interconnect design
> >
> > Hello,
> >
> > Ansofts HFSS version 8 translates modes to nodes , in other words
> > you
> > can
> > get
> > a fully coulped SPICE model of the device of interest to a very
> > high
> > bandwidth.
> >
> > Thank you.
> >
> > **** To unsubscribe from si-list or si-list-digest: send e-mail to
> > [email protected]. In the BODY of message put: UNSUBSCRIBE
> > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> > si-list archives are accessible at http://www.qsl.net/wb6tpu
> > ****
>
> --
> Fred Balistreri
> [email protected]
>
> http://www.apsimtech.com
>
> **** To unsubscribe from si-list or si-list-digest: send e-mail to
> [email protected]. In the BODY of message put: UNSUBSCRIBE
> si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> si-list archives are accessible at http://www.qsl.net/wb6tpu
> ****
>
>

**** To unsubscribe from si-list or si-list-digest: send e-mail to
[email protected]. In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Wed Nov 22 2000 - 10:51:02 PST