[SI-LIST] Re: Package SI vs. PCB SI
- From: "Haller, Robert" <rhaller@xxxxxxxxxxxxx>
- To: <si-list@xxxxxxxxxxxxx>
- Date: Tue, 14 Feb 2006 12:31:06 -0500
Hassan, Giancarlo,
IMHO a hybrid SI process is required to handle the complexities of many
of today's high speed PCB's. Complex accurate and complete I/O and
interconnect models are required for high speed links, but do not ignore
packages, power distribution or medium and slow speed networks. Murphy
is always out there and whatever you fail to consider (analyze
adequately) will come back to bite you.
Regards,
Bob=20
Robert Haller
Enterasys Networks
Core Routing - Hardware
Phone: 978-684-1340=20
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
On Behalf Of Ing. Giancarlo Guida
Sent: Tuesday, February 14, 2006 11:48 AM
To: hassan@xxxxxxxx
Cc: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Package SI vs PCB SI
Hi Hassan,
to my opinion there are at least 2 good reason that explain your
statement.
SI simulation accuracy depends on 2 factor:
1) accuracy of the model for the interconections
2) accuracy of the model fo the I/O Buffer
it is quite clear that when the data-rates become critical it is=20
necessary to use a 3D full wave approach in the
interconnect models
I think that tools like the ones you mention does not have such 3D=20
solver because they are
dedicated to post-layout analysis of the entire PCB and the usage of an=20
heavy 3D solver would be beyond their application area.
May be a similar argument could be used to explain the need to use spice
model for the I/o buffer instead of the ibis model...but I am not an=20
expert in this area.
Regards
Giancarlo
Hassan O. Ali ha scritto:
> Scott,
>
> Thanks for this excellent response.
>
> I have a question for you SI folks:=20
>
> Am I right to observe that as the data-rates increase beyond 5Gbps,
there is an=20
> increasing loss of faith on the previously popular board-level SI
tools such as=20
> SpecctraQuest (now Allegro PCB SI), XTK, Hyperlynx, ICX, etc. in favor
of HSPICE used=20
> with field-solver-s-parameter-based inteconnect models?
>
> I was personally surprised to see a notable absence of a focus on such
board-level tools=20
> in the just ended DesignCon 2006. What does this all mean? Is
Darwinism (survival of the=20
> fittest) at play here?
>
> I'd appreciate your responses.
>
> Best regards.
>
> Hassan.
>
>
>
> On Feb 14, Scott McMorrow <scott@xxxxxxxxxxxxx> wrote:
> =20
>> Prasad
>> The theory is the same,whether you are dealing with packages, system
in=20
>> package, MCM or PCB. However, the details will change quite a bit.
PCB=20
>> SI tools tend to assume that all routing is orthogonal, or aligned at
>> increments of 45 degrees. This makes sense, since orthogonal and 45=20
>> degree diagonal routing are extremely efficient with respect to space
>> utilization. An assumption is therefore made by the PCB trace=20
>> extraction tools that two traces are either parallel, or they are
not. =20
>> If they are parallel (as defined by parallelism on the 8 compass=20
>> directions) then coupling is calculated. If not, no coupling is=20
>> calculated. For all other traces, with non-45 degree angles, or=20
>> different angles, they are assumed to be uncoupled and independent.
As=20
>> I said, for PCB designs, this is often a reasonable assumption. But
for=20
>> aggressive PCB design where odd-angle routing is used for maximum
trace=20
>> density between parts, and for MCMs and packages, where routing often
>> radiates from a die in the center out to the balls, PCB-type trace=20
>> modeling software can be woefully inadequate. That's where you start
>> getting into the problem of non-uniform coupled transmission lines
and=20
>> imperfect planes with large numbers of perforations.=20
>>
>> Current density on package power and ground planes is much higher
than=20
>> that on PCB planes, causing additional SSO, SSN problems that are not
as=20
>> likely to impact signals on a PCB. As a result, analysis of package
SI=20
>> usually involves a combination of 3D solvers, and complete simulation
of=20
>> the package signal, power, ground and return paths. Books like Eric=20
>> Bogatin's give you all the basic information that you need. You can=20
>> then expand your knowledge with Brian Young's book, Digital Signal=20
>> Integrity, which has a very good discussion of partial inductance and
>> package measurements, Clayton Paul's book, Analysis of Multiconductor
>> Transmission Lines, which is still the fundamental theoretical book
on=20
>> TEM transmission line modeling, and Luc Martens book, High-Frequency=20
>> Characterization of Electronic Packaging, which gives a good overview
of=20
>> modern package characterization techniques.
>>
>> At the IC level, SI becomes even more difficult. There are no ground
>> planes, only power and ground meshes. Multiple modes of propagation
can=20
>> exist, including slow-wave mode, where the electric and magnetic
fields=20
>> decouple, due to the semiconductor properties of the dielectric,
rapidly=20
>> increasing propagation delays, and causing significant dispersion.
Ed=20
>> Sayre, III, on this list, has some background in this area. I am
sure=20
>> there are others lurking on this list that also have significant=20
>> experience in on-chip SI.
>>
>> Best regards and happy learning,
>>
>> Scott
>>
>>
>> Scott McMorrow
>> Teraspeed Consulting Group LLC
>> 121 North River Drive
>> Narragansett, RI 02882
>> (401) 284-1827 Business
>> (401) 284-1840 Fax
>>
>> <a=20
>> =20
>
href=3D'http://www.teraspeed.com</a>'>http://www.teraspeed.com'>http://ww=
w
.teraspeed.com</a
> =20
>> Teraspeed(r) is the registered service mark of
>> Teraspeed Consulting Group LLC
>>
>>
>>
>> prasadsa wrote:
>> =20
>>> Thanks for all your mails. Many of these books talk about refers to
SI on=20
>>> Boards. But i am looking in particularly for books on IC-Package if
any or=20
>>> does the same theory (books) holds good on IC-Package also.
>>>
>>> Regards,
>>> Prasad S A
>>>
>>> ----- Original Message -----=20
>>> From: "Coe, Stephen" <scoe@xxxxxxxxxxxxx>
>>> To: <qazi@xxxxxxxxxx>; <si-list@xxxxxxxxxxxxx>
>>> Sent: Tuesday, February 14, 2006 2:54 AM
>>> Subject: [SI-LIST] Re: Good book
>>>
>>>
>>> =20
>>> =20
>>>> Let me add my 2 cents :=3D20
>>>>
>>>> "High-Speed Circuit Board Signal Integrity"=3D20
>>>>
>>>> by Stephen Thierauf
>>>>
>>>> Regards,
>>>>
>>>> Steve
>>>>
>>>> -----Original Message-----
>>>> From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx]
>>>> On Behalf Of Qazi Arif Iqbal
>>>> Sent: Monday, February 06, 2006 6:37 PM
>>>> To: si-list@xxxxxxxxxxxxx
>>>> Subject: [SI-LIST] Re: Good book
>>>>
>>>> Eric Bogatin's "Signal Integrity Simplified" is another excellent
book.
>>>>
>>>> Qazi=3D20
>>>>
>>>> -----Original Message-----
>>>> From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx]
>>>> On Behalf Of Doug Martens
>>>> Sent: Sunday, February 05, 2006 8:25 PM
>>>> To: akhilesh.chandra@xxxxxx; prasadsa@xxxxxxxxxx
>>>> Cc: si-list@xxxxxxxxxxxxx
>>>> Subject: [SI-LIST] Re: Good book
>>>>
>>>> Black Magic is written by Howard Johnston and it is a good book.
>>>> Another good book is from Lee Ritchey called "Right the first Time:
a
>>>> practical handbook on high speed PCB and System design"
>>>>
>>>> -----Original Message-----
>>>> From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx]
>>>> On Behalf Of Akhilesh CHANDRA
>>>> Sent: February 5, 2006 9:19 PM
>>>> To: prasadsa@xxxxxxxxxx
>>>> Cc: si-list@xxxxxxxxxxxxx
>>>> Subject: [SI-LIST] Re: Good book
>>>>
>>>> Hello Prasad,
>>>>
>>>> In my view for signel integrity best book is "BLACK MAGIC" I =3D
>>>> forgot=3D3D20
>>>> the name of the writer. I will serch it on net and inform you soon.
>>>>
>>>> Regards
>>>> Akhilesh
>>>>
>>>> prasadsa wrote:
>>>>
>>>> =20
>>>> =20
>>>>> Hi,
>>>>> Can you please suggest me a good book for IC-Packaging / SIP
Technology
>>>>> =20
>>>>> & SI book on the same subject.
>>>>>
>>>>> Thanks in advance
>>>>> Prasad
>>>>>
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>>>> =20
>>>>> =20
>>>> ++++++++++++++++++++++++++++++++++++
>>>> =20
>>>> =20
>>>>> This e-mail is intended for the use of the individual or entity
named
>>>>> =20
>>>>> =20
>>>> above and may contain information that is confidential and
>>>> privileged.=3D3D20
>>>> =20
>>>> =20
>>>>> If you are not the intended recipient, you are hereby notified
that any
>>>>> =20
>>>>> =20
>>>> dissemination, distribution or copying of this e-mail is strictly
>>>> prohibited.
>>>> =20
>>>> =20
>>>>> If you have received this e-mail in error, please notify us
immediately
>>>>> =20
>>>>> =20
>>>> at prasadsa@xxxxxxxxxx and destroy the original message.=3D3D20
>>>> =20
>>>> =20
>>>>> While this e-mail and any attachments have been scanned for common
>>>>> =20
>>>>> =20
>>>> computer viruses and found to be virus free,=3D3D20
>>>> =20
>>>> =20
>>>>> we recommend you also perform your own virus checking processes
before
>>>>> =20
>>>>> =20
>>>> opening any attachments. =3D3D20
>>>> =20
>>>> =20
>>>>>
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>>>> =20
>>>>> =20
>>>> ++++++++++++++++++++++++++++++++++++
>>>> =20
>
> ------------------------------------------------------------------
> 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 FAQ wiki page is located at:
> http://si-list.org/wiki/wiki.pl?Si-List_FAQ
>
> List technical documents are available at:
> http://www.si-list.org
>
> List archives are viewable at: =20
> //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
> =20
>
>
>
> =20
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
List archives are viewable at: =20
//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
=20
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
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:
