RE: [SI-LIST] : the old high-frequency return current model

Eric Bogatin ([email protected])
Mon, 4 Oct 1999 12:16:48 -0500

Craig- Now that I no longer work for Ansoft, I can offer my own unbiased,
personal, professional opinions about their tools.

First, keep in mind that Ansoft offers about 20 different field solver type
tools, and they are all different and independent. Unfortunately, the
company has done a really poor job of name recognition for the individual
products. Most people lump all the field solvers in the general name of the
"ansoft" solver, or "Maxwell" solver. Neither of these names distinguish
any one of the different 20 products.

The particular field solver I was referring to in my past note on 2D field
solvers was the Maxwell 2D Extractor. It is a FEM based 2D solver. This is
distinctly different from the one I think you are referring to, a 2.5D
boundary element solver, outputting S parameter terms directly, which used
to be called Strata- has a new name now that I don't recall.

I am not really familiar with the boundary solver tool- the old Strata, and
can't comment on it- maybe someone else can. However, I am very familiar
with the 2D FEM solver, Maxwell 2D Extractor, part of the SI 2D suite. In my
professional opinion, "its a damn fine tool". I have verified the accuracy
to better than 1%- a few of the papers I wrote with others on this are on my
web site for download. I would love to get feedback from others on this


Eric Bogatin
Training for Signal Integrity and Interconnect Design
26235 W. 110th Terr.
Olathe, KS 66061
v: 913-393-1305
f: 913-393-1306
pager: 888-775-1138
e: [email protected]

> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of C Deibele
> Sent: Monday, October 04, 1999 11:24 AM
> To: [email protected]
> Subject: Re: [SI-LIST] : the old high-frequency return current model
> > Other field solvers, such as the Maxwell 2D Extractor, from Ansoft, will
> > solve LaPlace's equation to get the capacitance matrix elements, and
> > separately, Ampere's equation, to get the high frequency current
> > distribution, and then the magnetic fields, and then the
> inductance matrix
> > elements. The boundary condition it uses is that at high frequency, skin
> > depth is so small, there is no field inside the conductor and it behaves
> > like a perfect conductor, so the B field is only tangential at
> the surface,
> > no normal component. I have compared the two results- Hyperlynx's and
> > Ansoft's tool- and the agreement is within 2% across a wide
> range of aspect
> > ratios.
> >
> > In addition, the Ansoft tool, for example, will separately, if
> requested,
> > solve the Helmholtz equation (frequency domain solution for J and B) and
> > extract the current distribution inside the conductor (using FEM), at a
> > specified frequency. This will allow you to see the actual current
> > distribution inside the conductor, as frequency is varied. The
> plots I sent
> > around a few weeks ago were created using this tool. That's why
> the Ansoft
> > tool is often referred to as the "Rolls Royce of field solvers".
> >
> I actually had lots of problems with Ansoft's program. If phase is
> of any importance, they have a lousy program for calculating it. I have
> been working with Sonnet software -- they have a program like Ansoft's
> and it works great. It isn't perfect, mind you, but it is much much
> better.
> I don't know about the cost differences though. It all boils down to
> what the particular program uses for the MoM as green's functions.
> Ansoft
> is sort of new to the business of MoM for enclosed structures. I think
> in awhile they will have a more mature product. Not yet though.
> Definitely try out sonnet's program. i believe.
> Craig
> **** To unsubscribe from si-list: send e-mail to
> [email protected] In the BODY of message put:
> UNSUBSCRIBE si-list, for more help, put HELP. si-list archives
> are accessible at ****

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