[SI-LIST] Re: Transmission line simulation at 1.25GHZ
- From: Rob Hinz <rob@xxxxxxxxxx>
- To: si-list@xxxxxxxxxxxxx
- Date: Tue, 24 Jul 2001 13:07:30 -0700
Ashok,
In the work we have done here at SiQual on OC48 and OC192 systems we follow
this general procedure.
First we use a field solver to develop good, coupled line SPICE models for
the transmission system. Generally, we will use a combination of AnSoft 2D
and 3D field solvers and the HSPICE field solver for this purpose. We
believe using a field solver is required because it is the only way to
accurately model the properties of the transmission system, in particular
the impedance, coupling, and the loss due to skin effect and the
dielectric, both of which become significant at these frequencies. The 2D
solver is used primarily to produce the lossy, coupled line models and the
3D solver for things like vias that can be modeled reasonably well as
lumped elements, at these frequencies. We use the information the fields
solvers provide to produce a SPICE model for the transmission system. We
use HSPICE preferentially. So for us, this will usually include several
segments of coupled W-elements with lumped elements at various locations to
model vias and terminations.
Regarding the actual analysis, you may be able to infer what you need from
rise time and AC sweep but we find that it is usually a better and more
direct approach to simulate the interconnect scheme as part of the overall
system. That is, we will simulate the driver to the receiver using multiple
aggressors against a victim. We set up HSPICE so that we can see and
measure the data eye pattern. We use a bit pattern that will push the
system into all the data dependency corners as well as making certain we
have hit the process corners for both the silicon and the substrate. To
accomplish this you will need good device models from your silicon vendors
as well as a clear understanding of the process tolerances your substrate
suppliers can hold.
As you develop your methodologies, it is always helpful, when possible, to
compare your simulation results to actual measurements. This will allow you
to tune your process and gain confidence in your approach.
Rob Hinz
Senior Consulting Engineer
SiQual, Signal Quality Engineering
18735 SW Boones Ferry Road
Tualatin, OR 97062-3090
(503) 885-1231
http://www.siqual.com
At 12:43 PM 7/24/2001 +0530, you wrote:
>Hi all,
>
>I am supposed to do the simulation of transmission lines of OC48 section.
>Since I am a newby to this, I dont know where to start,, what are the things
>needed for this, etc. The rough idea I have is to simulate it in SPICE
>(WinSPICE or SPICE OPUS). I would like to know,
>
>1. Which SPICE transmission line model is to be used? How to get deduce its
>parameters from the PCB line parameters for the same? How differential lines
>are modelled in SPICE?
>
>2. What are the analysis to be done? Will transient analysis (for rise time
>degradition etc) and ac Sweep analysis would suffice?
>
>The very idea I am having may be incorrect. Please guide me in this work.
>
>Thanks and Regards,
>Ashok.
Rob Hinz
SiQual Corporation
rob@xxxxxxxxxx
phone (503)885-1231 x30
fax (503)885-0550
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List archives are viewable at:
//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:
