RE: [SI-LIST] : SI tool for signals above 1GHz

About this list Date view Thread view Subject view Author view

From: WANG,LIN-WL (A-Singapore,ex3) ([email protected])
Date: Thu Feb 01 2001 - 18:39:43 PST


Carlos:

Your suggestion is helpful to me. Thanks! About simulation, I have some
questions shown below:

-----Original Message-----
Wang Lin:

Maybe some useful information.....

Form a software point of view, a signal with a bandwidth of 2.3Ghz is well
within the limitations of the field solutions generated by solvers (mostly
bem) employed by most SI software companies.

However, the question you present is not imposing on the ability of the
field solvers to generate accurate solution at high frequencies, but with
todays models and tools to properly use these models and . In order to make
an intelligent decision on which vendor to use it is my recommendation to
look at the entire process;

1. What do you want to do with your tool?
        What type of models will it support? IBIS (Behavioral I/O buffer
models, limited to low Ghz)? SPICE? Macromodels? etc..

I've heard an opinion that IBIS model is only good below 1GHz, but SPICE
model can still be used in GHz range. Can you recommend any document
clarifying this point in detail?

        What capabilities do you require for integration with other
software, including utilities to convert proprietary single line/coupled
line elements to other vendors? 3D field solvers? RF simulators?
        Do you need parametric sweep capabilities for Solution Space
Analysis?
        Is there a need for SI simulation on pre routed designs or post
routes, which may include DIE, Package and PCB?

Definitely need the pre-layout analysis. The question is what problems
should be considered at this level. Of course layer structure, trace
impedance, termination, crosstalk for parallel traces need to be considered,
but I'm afraid these are not enough. Do you have some supplement? Same
question apply to post-layout analysis. And for how fast signal, the
packaging model must be considered in the simulation?

        Controlled current and current sources (simple pulse, PWL,
exponential, sinusoidal, etc?
        Custom stimulus? Jitter? Customization for special measurements
(source synchronous vs. common clock, etc.)?
2. How is the integration to Silicon? Package? PCB?
3. Plane resonance? Proper decoupling capacitor placement and analysis?
4. Can a tool accommodate for conductor losses(skin) for the entire process,
from Silicon/Die to Package and PCB?
5. Can you run the tool in batch mode?
6. Can the tool support timing information to compensate for margins (i.e,
tco, setup and hold, etc?)
7. Is there report capabilities to easily format large simulation runs?
8. How good is the waveform viewer and what capabilities are offered within
the viewer (this is important)?
9. What customer base is using the tool?
10. How stable is the company in this market place?
etc.....

Again, your process and requirements will determine the tool you need. But
I believe looking at the big picture (process) can be an effective way for
paving the groundwork for a sound decision.

Carlos Moll
Cadence Design Systems, Inc.

At 05:38 PM 1/31/01 +0800, WANG,LIN-WL (A-Singapore,ex3) wrote:
>Hello,
>
>I'm trying to do SI analysis for differential signals with clock rate above
>1GHz, rise time around 150ps. I'd like to ask whom already have successful
>SI experience in this frequency range that:
>1) what software tool you recommend to do SI?
>2) any important issue should be noted during the analysis?
>3) any other good suggestions?
>
>For the software, I ever considered XTK, but not sure if it can be used at
>this freq range? Another one in my mind is ADS1.3, does anyone ever use it
>for SI?
>
>Thanks!
>
>Wang Lin
>Agilent, Singapore
>
>**** To unsubscribe from si-list or si-list-digest: send e-mail to
>[email protected]. In the BODY of message put: UNSUBSCRIBE
>si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
>si-list archives are accessible at http://www.qsl.net/wb6tpu
<http://www.qsl.net/wb6tpu>
>****

**** To unsubscribe from si-list or si-list-digest: send e-mail to
[email protected]. In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Tue May 08 2001 - 14:30:47 PDT