[SI-LIST] Re: Preparing S-Parameters for Simulation

  • From: <amolak_badesha@xxxxxxxxxxx>
  • To: <shlepnev@xxxxxxxxxxxxx>, <tim@xxxxxxxxxxxxxxxx>, <si-list@xxxxxxxxxxxxx>
  • Date: Wed, 25 Feb 2009 12:39:45 -0700

Hi Yuriy and Timothy, 

The reason typical Spice tools have tough time with S-parameter models is 
because they do not have comprehensive causality and passivity 
checks/corrections built into the tool. However, there are Spice tools 
available now (like Agilent ADS), that can accurately simulate S-parameters 
directly in Time-domain. This is done with built-in causality/passivity 

For many customers, converting S-parameters models to broadband spice models is 
an extra step which can be avoided if the Time-domain Spice tools do the model 
"correction" directly. Also, there is a question of the fidelity of the 
broadband spice model - the accuracy depends on how complicated the channel 
response is and how many poles/zeros are used. 


Amolak S. Badesha
Agilent EEsof EDA

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On 
Behalf Of Yuriy Shlepnev
Sent: Wednesday, February 25, 2009 7:15 AM
To: 'Timothy Coyle'; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Preparing S-Parameters for Simulation

Hi Timothy,

Though, this topic was discussed here multiple times, I think it worth to
revisit it from time to time, considering the recent developments in this
area. Here are my takes on it.

First of all, you have to distinguish two S-parameters use cases during the
system-level analysis. 
S-parameters may be used directly in discrete or tabulated form with
possible interpolation in frequency domain and with some IFFT or convolution
algorithms in time domain.
Alternatively, S-parameters can be converted into rational compact
macro-model both for frequency and time-domain analysis. The rational
macro-models usually provide superior accuracy and performance both in time
and frequency domain analyses and ideally should be used as the models of
multiport structures (in form of broad-band SPICE models for instance)
whenever possible instead of the Touchstone tabulated models.

Considering setting the bandwidth for electromagnetic analysis, the low
frequency should be below the transition to skin-effect (lower MHz for PCBs)
to allow safe extrapolation to DC. Note, that this is not possible with
tools that use SIBC to simulate the skin-effect. Typically, the high
frequency should be defined by the required resolution in time-domain
(1/2tr). Though that may be not possible due to propagating high-order modes
and this frequency may be set to a smaller value if your solver builds
rational compact macro-model with the delay extraction - it technically
provides the extrapolation. The resolution and spacing are very important
for IFFT and convolution algorithms and were discussed here before, but not
that important to build rational macro-models, as soon as there are 3-4
frequency points for each resonance.

Considering the quality analysis and restoration, I would recommend to study
this paper:
P. Triverio, S. Grivet-Talocia, M.S. Nakhla, F.G. Canavero, R. Achar,
Stability, causality, and passivity in electrical interconnect models, IEEE
Trans. on Adv. Pack., v. 30, N4, p. 795-808.
The best way to compare the original and "restored" S-parameters is to plot
and compare them in frequency domain (phase and magnitude). The original and
restored S-parameters are typically almost identical in case of small
violations and visually different in case of large violations. The data
should be discarded in the later case as rightly pointed out in the cited

Considering the quality of S-parameters from the electromagnetic analysis,
violation of reciprocity and passivity are relatively rare (only in case of
algorithm failure). Properties of S-parameters due to geometric symmetry can
be violated with non-symmetric meshing. The causality is often the issue due
to non-causal models of dielectrics and conductors used in some
electromagnetic analysis tools. This is not a problem for narrow-band
microwave applications, but a problem for the broad-band digital
application. Causality of such S-parameters can be restored by building the
rational macro-model, but it does not actually improve the quality of the
model and the model typically has to be discarded and rebuilt with the
broad-band causal models both for dielectrics and conductors.

Best regards,
Yuriy Shlepnev

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On
Behalf Of Timothy Coyle
Sent: Tuesday, February 24, 2009 9:52 AM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Preparing S-Parameters for Simulation

I am working on a tutorial for the next issue of XrossTalk Magazine about
getting S-Parameter models ready for simulation. I see a lot of S-Parameter
models from lab measurements and simulation tools that have a lot of quality
issues (causality and passivity) and have to do a lot of clean up to make
them suitable for simulation. I have my own bag of tricks I use as well as
some good books on S-Parameter theory and application notes but I would like
to hear from other people what their approaches are or some good resources. 


For some particular specifics review the list below:

.         What considerations should go into setting up a simulation tool to
extract an S-Parameter model? (bandwidth to be used, number of points,
importance of linear spacing of points, starting at DC, etc) 

.         A lot of tools will automatically "fix" causal and passivity
issues or enforce them. What's the pros and cons for this? What do engineers
need to look out for when they do this?

.         What's the best way to clean up a lab measured S-Parameter model
with poor resolution?

.         What type of quality checks should you perform on an S-Parameter
model before simulating? (what do you look for in the S21 plots, do you use
smith charts, ect?)


Thanks for any and all suggestions and links. 




Timothy Coyle

Editor In Chief

XrossTalk Magazine

405 Western Ave #430

South Portland, ME  04106

Tel:  617.297.2566

Fax: 207.510.8099

Email: tim@xxxxxxxxxxxxxxxx

http://www.xrosstalkmag.com <http://www.xrosstalkmag.com/> 

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:

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List technical documents are available at:

List archives are viewable at:     
or at our remote archives:
Old (prior to June 6, 2001) list archives are viewable at:

Other related posts: