[SI-LIST] Re: Regarding IBIS pkg modelling

Hi Amit
the accuracy of your PKG extraction is related to mesh accuracy.

If you are using quasi-static field solver (like Q3D), set the mesh at
highest frequency (3*Fknee) and then export RLCG at fundamental frequency
(for better accuracy you can export multiple cascaded cells instead of
single RLCG pi-greco model)

If you are using Full-Wave-Solver (like HFSS), set the freq. mesh at Fknee
(or Fmax) and then get S-matrix from DC up to Fmax

The accuracy of field solver is strictly related to mesh accuracy. REMEMBER
it's all a matter of mesh accuracy
If you set properly this parameters, you will get accurate model

BTW some EDA tools (like Q3D, HFSS etc.) have automatically mesh adapter
algorithm that adjust the mesh, during loop cycle,
based on local field error on each mesh element --> this means that you will
have high dense mesh where there is a strong field energy, and coarse one
where the field is low

This will allow you to get very high accuracy!

Regards
Cris

2011/7/8 Amit KUMAR STE <amit.kumar@xxxxxxxxxxxxxx>

> Hello Cristian ,****
>
> ** **
>
> My doubt is that if we extract our pkg at this third harmonics , the pkg
> file is giving us pessimistic value for the fundamental frequency of the
> signal.****
>
> That means pkg file is not correct. Am I right?****
>
> ** **
>
> Regarding s parameters I am asking that suppose I have extracted s
> parameter values for upto 2GHz and I am using it for DDR simulation with
> freq of 533 MHz.****
>
> Which values the simulator tool will consider for this simulation? will it
> take the s parameter values which is present for 533 MHz or will it take
> values of 533MHz along with values of 1066 and 1599 MHz also and somehow
> come up with the values which is most suitable for this simulation?****
>
> ** **
>
> Hope I am clear this time J****
>
> ** **
>
> Regards****
>
> Amit****
>
> ** **
>
> ** **
>
> ** **
>
> *From:* Cristian Gozzi [mailto:cristian.gozzi@xxxxxxxxx]
> *Sent:* Friday, July 08, 2011 12:11 PM
> *To:* Amit KUMAR STE
> *Cc:* si-list@xxxxxxxxxxxxx
>
> *Subject:* Re: [SI-LIST] Regarding IBIS pkg modelling****
>
> ** **
>
> Hi Amit
>
>
> when you get a PKG model at one freq. means you are getting a lumped RLCG
> model
>
> In this case the PKG was simulated at knee freq. or sometimes also at
> 3*Fknee (Fknee=0,5/Trise) in order to get the best accuracy also for highest
> digital harmonics
> (remember: lumped model is accurate until lenght of nets/PKG are less than
> lambda/10)
>
> Instead if you get S-parameters, this model is usually broadband model,
> from DC up to Fmax=3*Fknee
>
> S-parameters capture all electromagnetic phenomenon (including radiation
> losses, copper roughness etc.)
>
> hope this help
> Regards
> Cris****
>
> 2011/7/8 Amit KUMAR STE <amit.kumar@xxxxxxxxxxxxxx>****
>
> Hello All ,
>
> When we get any package model extracted , it gives us the coupling values
> at certain frequency.
> A digital signal always has many harmonics of a fundamental frequency and
> hence all those frequency components are important.
> Which frequency is appropriate for package model extraction then?
>
> Also if we use s-parameter model , does the simulator tool take into
> account the s parameter values at multiple frequencies and come up with the
> appropriate coupling values ?
>
>
> Regards
> Amit
> ------------------------------------------------------------------
> 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 technical documents are available at:
>                http://www.si-list.net
>
> List archives are viewable at:
>                http://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:
http://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:     
                http://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: