[SI-LIST] Re: Current Return Vias

Istvan,
I will agree, sort of.  Design for clean power, signal return path 
integrity, and emissions are all important design tasks, each affecting 
the other.  I like your unified approach.  As you know, Steve and I 
think along similar lines.  However, in general usage in the literature, 
on the si-list, and in the minds of most readers, PDS generally 
encompasses the power delivery aspects of a design from VRM to die, 
while excluding the signal paths, which are generally known as SI 
related issues.  But, as we know, there are intersections and overlaps 
between each of these.

Since most readers treat PDS design as a separate issue from signal path 
design, I think it's appropriate to state that a PDS only design 
approach to noise reduction may not be enough to solve all problems. 

regards,

Scott


Scott McMorrow
Teraspeed Consulting Group LLC
121 North River Drive
Narragansett, RI 02882
(401) 284-1827 Business
(401) 284-1840 Fax

http://www.teraspeed.com

Teraspeed® is the registered service mark of
Teraspeed Consulting Group LLC



Istvan Novak - Board Design Technology wrote:
> Scott,
> I think it depends on what one considers to be a good PDS design.
>
> In my view, the PDS functions are:
>
> - provide clean power to the chips (primary function)
> - provide adequate reference path for signals (optional function)
> - make sure that the radiation is within allowed limits (optional function)
>
> The second and third functions are optional, because it depends on the
> construction whether those are functions of the PDS that we depend on
> or not.  The design target is not necessarily the same along those
> three functions: a design, which is good enough to provide sufficiently
> clean power to the silicon, may fail in providing good enough return
> path or may emit above allowed limits, and vica versa.
>
> Regards,
> istvan
>
>
>
> Scott McMorrow wrote:
>
>   
>> Ding, ding, ding ... this is a good example of a design with a poor 
>> return path due to signal via crowding.  Adding ground vias within the 
>> signal via field, for a gnd referenced design, will undoubtedly help, as 
>> long as the ground vias are interspersed within the signal via field.  
>> The PDS will not help you when signal edge rates are in the several 
>> hundred picosecond range.
>>
>> Those who believe that "good" PDS design will solve all signal via 
>> related crosstalk problems are woefully mistaken.
>>
>>
>>
>>
>> Scott McMorrow
>> Teraspeed Consulting Group LLC
>> 121 North River Drive
>> Narragansett, RI 02882
>> (401) 284-1827 Business
>> (401) 284-1840 Fax
>>
>> http://www.teraspeed.com
>>
>> Teraspeed® is the registered service mark of
>> Teraspeed Consulting Group LLC
>>
>>
>>
>> Perry Qu wrote:
>>  
>>
>>     
>>> Interesting topic. We recently had a design with a SDRAM interface consists
>>> of 5 SDRAMs, 2 pairs clamshelled and one standalone. It's running on 155 MHz
>>> clock single data rate so not a fast interface by today's standard. For the
>>> clamshelled SDRAMs, we measured the address signals and the waveform on the
>>> top and bottom sometimes are quite different, specifically, one of the SDRAM
>>> (could be either top or bottom depending on which address pin we probe)
>>> shows very visible crosstalk and worse shelf on edge.  The result is less
>>> margin for timing. 
>>>
>>> This can't be caused by topology as we length-matched the trace from SDRAM
>>> pin to the branch via in a mixture of star/far-end cluster topology. The
>>> only thing I can think of is return current as most of the signals vias are
>>> under the middle of the SDRAM packaging with no ground/power via really
>>> closeby. The SDRAM packaging is not well done in my opinion as they have
>>> most of the power/ground pins on one half of the chip, and very few
>>> reference pins on the other half where all address pins locate. To make it
>>> worse,  the routing layer is all over the stackup and it can jump from a
>>> layer close to top to a layer close to bottom and then jump back, with no
>>> return vias close to signal vias, although they all reference to GND (we
>>> have multiple GND planes throughout the stackup). The worse waveform seems
>>> always occurs at the SDRAM with bigger jump in terms of reference planes
>>> thus larger return loop area, which is the cause for crosstalk and higher
>>> inductance on return and shelf on edge. I decide to add some ground vias
>>> close to the signal vias for our next design which uses same chips. 
>>>
>>> Regards
>>>
>>> Perry
>>>
>>>
>>> ======================================= 
>>>
>>> Perry Qu 
>>>
>>> Design & Qualification, Alcatel Canada 
>>>
>>> 600 March Road, Ottawa ON, K2K 2E6 
>>>
>>> DID: 613-7846720  Fax: 613-5993642 
>>>
>>> Email: perry.qu@xxxxxxxxxxx 
>>>
>>> ======================================= 
>>>
>>>
>>> ------------------------------------------------------------------
>>> 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 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:     
>>>             http://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
>>>  
>>>
>>>  
>>>    
>>>
>>>       
>> ------------------------------------------------------------------
>> 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 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:     
>>              http://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
>>  
>>
>>  
>>
>>     
>
> ------------------------------------------------------------------
> 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 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:     
>               http://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
>   
>
>   


------------------------------------------------------------------
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 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:     
                http://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: