[SI-LIST] DDR3 SI and Timing analysis
- From: Cristian Filip <cris_filip2002@xxxxxxxx>
- To: si-list@xxxxxxxxxxxxx
- Date: Wed, 25 Jun 2008 19:51:23 -0700 (PDT)
DDR3 SI and Timing analysis
Hello,
I am working on a DDR3 interface and the final outcome should be the comple=
te SI and Timing analysis. We are using Mentor tools (HyperLynx, ICX Pro & =
ICX) and I am just at the beginning of this process. I am sure that until t=
he end I will have more questions on this topic and this is why I preferred=
to give the actual title to this message and not a more specific one. I am=
sure also that other people on this group are facing similar situations, s=
o we might exchange some of our experiences.
The supplier of the Memory Controller has provided the IBIS model of the di=
e and four Package Electrical models in HSPICE format extracted using Ansof=
t Q3D Extractor. We got also a schematic diagram which specifies the order =
(serial) we have to follow to connect those models.=20
On the memory side we are going to use SODIMM=E2=80=99s and we got the EBD =
and IBIS models for those DIMM=E2=80=99s from our supplier.=20
I have tried to do some pre-layout simulations in LineSim using the models =
provided and on a very simple point to point DQ topology I got a very ugly =
signal at the DIMMs, which seem to be determined by a huge attenuation intr=
oduced by the package parasitic, so I decided to validate my model before t=
o investigate the spice models.=20
From the IBIS model I determined that Vref =3D1.5V, Rref =3D50Ohms and Cref=
=3D0. Many documents I have read are suggesting that the test load would be=
in this case a simple pull-up resistor connected to a 1.5V power rail. I d=
id try and the results of simulation are showing a DC offset, so the Vol DC=
=3D600mV > Vil DC=3D595Mv, which tells me that the test circuit isn=E2=80=
=99t the right one. If I replace the Vref with 0.75V the output signal make=
s sense.=20
The questions that arise are:=20
=E2=80=A2Are the Vref, Rref and Cref the parameters describing the test loa=
d?
=E2=80=A2If yes, this holds true even if the IBIS model is actually the mod=
el of the die?
=E2=80=A2If we add to the IBIS models the series Spice parasitic models the=
test load will still be described by those parameters?
=E2=80=A2What would be the best way of using this combination of IBIS-HSPIC=
E models in LineSim and BoardSim?
=E2=80=A2It is possible to run batch mode simulation in BoardSim using this=
combination of models?
=E2=80=A2If in top of those we add some timing models what would be the com=
bined solution to do SI and timing analysis?
=E2=80=A2It is ICX or ICX Pro (Explorer and Verify) more suitable for this =
kind of situation?
Thank you in advance,
Cristian Filip
=0A=0A=0A ____________________________________________________________=
______=0AYahoo! Canada Toolbar: Search from anywhere on the web, and bookma=
rk your favourite sites. Download it now at=0Ahttp://ca.toolbar.yahoo.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:
//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
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:
