Re: [SI-LIST] : skin effect and transmission line simulators

Al Barber (barber@hplms2.hpl.hp.com)
Thu, 04 Sep 1997 11:47:07 -0700

Chris Simon wrote:
>

> The simulator I'm looking for should include all of the following:
....
> 2. Transmission line models that include series resistance and
> inductance which are functions of frequency to ACCURATELY model skin
> effect. (i.e., a frequency dependent line model that is used to get time
> domain simulation results)

MDS does accurately model skin effect and does include frequency
dependent inductance. At least you can watch the Z0 of a line change due
to the effect on internal inductance as current crowds to the surfaces.
But this is a relatively low frequency effect and I'm guessing not that
important to your application. It also models dielectric loss, though
not a frequency dependant loss tangent.

> 3. SPICE CMOS models and/or IBIS type non-linear models can be used
> as drivers and receivers

MDS handles IBIS non-linear models. It also includes a spice deck
translator, but it's my understanding (I don't speak for HP on this...)
that it is limited in its CMOS model capability.
....

>
> There is one other criteria, but I'm not certain that this one is required
> for signals with 150 to 200 pSec rise and fall times. Any opinions?
> 5. Includes the effect of dielectric losses.
>
Depends on how much dielectric loss you expect. Skin effect resistance
increases with the square root of frequency and dielectric loss
increases directly with frequency. So for any particular combination of
geometry, resistivity, and dielectric loss tangent there will be a
frequency above which dielectric loss dominates. For pcb trace
geometries and loss tangent around .03 (FR4 can be this bad) this
crossover can be as low as 1 GHz, which will be very noticeable at your
risetimes. For the higher quality dielectrics of cables or other pcb
dielectrics the frequency will be higher. If your simulator cannot
handle dielectric loss, your application flexibility will be limited.

....

> I would be interested in hearing which of the above criteria are met by
> other tools, or whether a future revision will incorporate these >features.

There is an additional feature I'm surprised you don't mention. How
about crosstalk among multiple lines?

Al Barber
Hewlett Packard Labs