[SI-LIST] Re: Simple Rational Model Estimation Method Given Transfer Function
- From: "Yuriy Shlepnev" <shlepnev@xxxxxxxxxxxxx>
- To: "'Charlie Sweet'" <csweet@xxxxxxxxxxxxxxx>
- Date: Thu, 8 Apr 2010 09:27:34 -0700
Hi Charlie,
Transmission parameter S21 is the wave transfer function for a 2-port system
terminated with the normalization impedances of the corresponding
S-parameters. No assumption about the reflection is required to fit it.
Depending on your application, the other options would be impedance or
admittance transfer functions. Fitting just transfer function is much
simpler task comparing to fitting a complete scattering or immitance matrix
for instance. The smoothness of the wave transfer function depends on the
t-line characteristic impedance and presence of the discontinuities in the
system. In context of signal integrity, a connection of 50-Ohm source and
load with 50-Ohm transmission line for instance will have zero reflection
loss the ideally smooth transfer function. Mismatches and discontinuities
lead to oscillations or resonances in the transfer function caused by the
reflections. In both cases it can be approximated with the rational basis
functions following the vectfit algorithm (may be without passivity
enforcement as in your case). Note, that it works only for linear
time-invariant systems.
Best regards,
Yuriy Shlepnev
www.simberian.com
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On
Behalf Of Charlie Sweet
Sent: Wednesday, April 07, 2010 8:37 AM
To: shlepnev@xxxxxxxxxxxxx
Cc: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Simple Rational Model Estimation Method Given
Transfer Function
Hi Yuriy,
Thanks for your feedback, I looked into the vectfit Matlab / Octave script
and found it focuses on fits to Y or S parameter matrices.
I am interested in pole-zero fits to a transfer function, not full 2 port Y
or S parameter matrices.
Can I assume S11=S22=0 and S12=S21 with magnitude of S21 = transfer function
gain and phase of S21 = transfer function phase to obtain a reasonable
pole-zero fit or is more required to obtain a causal, stable model? The
transfer functions typically are smooth functions with no discontinuities in
magnitude or phase versus frequency. The actual networks are not passive
and I am primarily interested in the forward direction, not reflections.
Thanks,
Charlie
On Wed, Mar 31, 2010 at 2:52 PM, Yuriy Shlepnev
<shlepnev@xxxxxxxxxxxxx>wrote:
> Charlie,
>
> VECTFIT is the software for rational macro-modeling developed by Bjorn
> Gustavsen - available at
> http://www.energy.sintef.no/Produkt/VECTFIT/downloads.htm
> Pretty much all fitting or rational macro-modeling tools in the industry
> are
> based on the vectfit algorithm (though the performance and capabilities
are
> very different in each implementation).
>
> Our free tool Simbeor 2008.01 L0 can be also used to build rational
> macro-models for 2-port structures (available at
> www.simberian.com/Downloads.php - item 2, the download requires
> registration
> at our web site).
>
>
> Best regards,
> Yuriy Shlepnev
> www.simberian.com
>
>
> -----Original Message-----
> From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
> On
> Behalf Of eenville
> Sent: Wednesday, March 31, 2010 11:31 AM
> To: si-list@xxxxxxxxxxxxx
> Subject: [SI-LIST] Simple Rational Model Estimation Method Given Transfer
> Function
>
> Any suggestions appreciated on simple, preferably open-source, tools to
> develop a good (e.g., in a least-squares sense for both real and imaginary
> components) low order rational model (i.e., pole/zero) given magnitude and
> phase of a transfer function vs. frequency? The phase info is relatively
> smooth but the roll-off and passband do not look like a second or even 3rd
> order model would suffice.
>
> Any ideas appreciated!
>
> Thanks,
>
> Charlie
>
> ------------------------------------------------------------------
> 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:
> //www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
>
> List technical documents are available at:
> http://www.si-list.net
>
> List archives are viewable at:
> //www.freelists.org/archives/si-list
>
> Old (prior to June 6, 2001) list archives are viewable at:
> http://www.qsl.net/wb6tpu
>
>
>
--
- Charlie
----
Charlie Sweet | Principal Engineer
Sunrise Labs | 5 Dartmouth Drive | Auburn | NH | 03032
Tel: 603-644-4500 | Fax: 603-622-9797 |
csweet@xxxxxxxxxxxxxxx | http://www.sunriselabs.com
New Product Development "from mind to market"
------------------------------------------------------------------
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:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
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:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
