[SI-LIST] Re: selection of Tx line models depending on frequency

Elen Parker Wrote:
>
> I have questions about until what frequency I can use "lossless
transmission
> line model" which is supported by XNS/TLC tools at XTK.
>
> I've heard that after the boundary frequency I had better use W-element
> whose characteristics are obtained by XFX Field Solver or HSPCIE Field
> Sovler. But I have a little idea until what freqeuncy XTK or Spectrqwest
can
> give the reliable simulation results like up to f=1 GHz ( DDR data rate)
or
> f=500 MHz(DDR data rate).
>
> If XTK cannot give the reliable results after 1 GHz(just example), is
HSPICE
> only tools which I can resort to?
> Thanks in advance.
>
> Elen

Dear Elen and others:

When appraising use of lossless vs. lossy transmission line models, in
addition to frequency there are several parameters which deserve
consideration. These include the signal edge rate, conductor characteristics
( trace length, width, thickness, resistivity, surface roughness) and
 various substrate properties (e.g. dielectric permitivity , loss tangent,
etc. ).

XTK is regarded suitable for simulation of digital designs
up to frequency of about 10 GHZ.  For XTK simulation of lossy lines it is
often necessary to satisfy the following two requirements:

1. In the Global Control File (.gcf) , accurate values must be specified for
RHO (Ohmic resistivity of signal and plane conductors),
TAN (dielectric loss tangent), LAMBDA (surface roughness factor),
and SIGMA (dielectric DC conductivity).

Furthermore, the relative dielectric constant ( EPS ) for substrate is
normally specified  in .gcf Stackup section  and a value for the
substrate relative Permeability ( MU ) can also be assigned in the Defaults
portion of .gcf file

2. In the XNS GUI the W-element loss algorithm should be enabled.


With RHO, SIGMA, TAN, etc. defined in .gcf file,
a .xfx file can be produced (by running ISF2XTK) having
loss parameters. Subsequently, execution of the XFX program outputs
a Transmission Line Parameters (.tlp ) file which contains loss matrices.

It should be added that while simulations which utilize lossy
models/techniques can produce more accurate results, this is often
accomplished at the expense of simulation speed.

The final paragraph depicts the prime features of LAMBDA .

It is a function of smoothness/roughness (e.g. tooth structure) of trace
surface and therefore, dependent on the manufacturing processes.
It is a dimensionless, non-negative parameter which forces XTK to use
multiple values of RHO for computation of skin effect resistance Rs
(but not for Ohmic resistance Rdc calculation).  Lambda can be determined
from the ratio of measured Rs ( utilizing a test coupon ) and Rs outputted
 by XFX.  The default value is 1.0 (representing a normal smooth surface  )
 which is quite satisfactory for frequencies below 200 MHz.
In a contrast with XTK, HSPICE does not require lambda for computing Rs.

I  wish all of you a very joyful holiday season and a prosperous New Year.

Abe Riazi
ServerWorks






------------------------------------------------------------------
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: