[SI-LIST] Re: One stitching via or more vias is better for 25Gbps application???

  • From: Bipin Dhavale <bipinpd@xxxxxxxxx>
  • To: "bertsimonovich@xxxxxxxxxx" <bertsimonovich@xxxxxxxxxx>
  • Date: Mon, 8 Sep 2014 08:56:44 -0700

In addition to paper you sent, below is one study with asymmetric differential 
pairs. ( most large bga breakouts wont have symmetric diff pairs)

http://www.univ-brest.fr/SPI/pages/Slides/SPI_2009_S7_1.pdf

Its not the exact same case of transition via's but still.



> On Sep 8, 2014, at 4:54 AM, Bert Simonovich <bertsimonovich@xxxxxxxxxx> wrote:
> 
> Here is a DesignCon2013 paper on the effect of nearby gnd via placement on HS 
> SE and Diff Signals by:
> A. Jaze; B. Archambeault; and S. Connor. 
> 
> http://e.ubmelectronics.com/images/DesignCon2014/12-WA2_Paper_EffectofNearbyGroundVia.pdf
> 
> Bert Simonovich
> 
> 
>> On Sep 8, 2014, at 3:35 AM, Leeyuyun <lee.home.61@xxxxxxxxx> wrote:
>> 
>> Hi, Steve
>> Could you help explain
>>> I think that you are considering a high frequency problem in low 
>>> frequency terms
>> I do not understand "in low frequency terms", could you help clarify it?
>> How to roughly evaluate cavity effect when stitching via close to signal 
>> vias?
>> Any white paper or material to show stitching via effect?
>> When I ask two parties to run two kinds of simulation, including only one 
>> stitching vias and more stitching vias, the result from two parties is 
>> different. One shows only one stitching is better, but another shows more 
>> stitching vias better. How do I make decision which one is correct?
>> Thank you
>> Lee
>> 
>>> steve weir <weirsi@xxxxxxxxxx> 於 2014/9/8 上午11:57 寫道:
>>> 
>>> It is neither.  It is a description of the physics.  Real engineering 
>>> accounts for pertinent physics.
>>> 
>>> Best Regards,
>>> 
>>> 
>>> Steve
>>>> On 9/7/2014 6:41 PM, leeritchey@xxxxxxxxxxxxx wrote:
>>>> This sounds like a bit of supposiition rather than real engineering.
>>>> 
>>>> Sent from my HTC One™ X, an AT&T 4G LTE smartphone
>>>> 
>>>> ----- Reply message -----
>>>> From: "steve weir" <weirsi@xxxxxxxxxx>
>>>> To: "Leeyuyun" <lee.home.61@xxxxxxxxx>
>>>> Cc: "si-list@xxxxxxxxxxxxx" <si-list@xxxxxxxxxxxxx>
>>>> Subject: [SI-LIST] Re: One stitching via or more vias is better for 25Gbps 
>>>> application???
>>>> Date: Sun, Sep 7, 2014 10:43 AM
>>>> 
>>>> 
>>>> I think that you are considering a high frequency problem in low 
>>>> frequency terms.  Stitch vias impact the cavity, and when they are in 
>>>> close proximity to signal vias and traces, they affect wave modes near 
>>>> those structures.  Thinking about stitch vias and the cavities that they 
>>>> traverse as though they are continuations of the TEM path that breaks at 
>>>> a Z axis transition is not a good approximation.  The Z axis signal 
>>>> transition injects the signal energy into the cavity.  The various 
>>>> properties of the cavity:  material, and geometry, including what vias 
>>>> stitch the cavity surfaces together and where affects how that signal 
>>>> energy propagates through the cavity in all directions.
>>>> 
>>>> Steve
>>>>> On 9/7/2014 5:31 AM, Leeyuyun wrote:
>>>>> Hi, Steve
>>>>> But IL and RL should depend upon the distance / via pad size of signal 
>>>>> via and gnd stitching via
>>>>> If only depend on capacitor vsd via, is the return path loop larger to 
>>>>> create large inductance?
>>>>> Does it fit into 25Gbps application?
>>>>> Could you advise what is the benefit of case 1 and case 2?
>>>>> Thanks
>>>>> Lee
>>>>> 
>>>>> 
>>>>>> steve weir <weirsi@xxxxxxxxxx> 於 2014/9/7 下午6:11 寫道:
>>>>>> 
>>>>>> The statement is based on the idea that Vss stitching from bypass
>>>>>> capacitor Vss vias and IC Vss vias is adequate.  The idea concludes that
>>>>>> explicit vias corresponding to signal or signal pair transitions are not
>>>>>> necessary.
>>>>>> 
>>>>>> Steve
>>>>>>> On 9/7/2014 2:46 AM, Leeyuyun wrote:
>>>>>>> Stitching should be needed when gnd reference plane change
>>>>>>> I do not know why you said do not need any stitching vias
>>>>>>> I want to clarify my question as below
>>>>>>> # means stitching vias, * means signal vias, + means positive signal, - 
>>>>>>> means negative signal
>>>>>>> 1. Only on stitching vias
>>>>>>> +++++++++++++++++*++++++++++
>>>>>>>                                  #
>>>>>>> --------------------------*----------------
>>>>>>> 2. At least two stitching vias
>>>>>>>                                 #
>>>>>>> ++++++++++++++++*+++++++++++
>>>>>>> 
>>>>>>> ------------------------*------------------
>>>>>>>                                 #
>>>>>>> For each structure, what is theory behind? What is your recommendation 
>>>>>>> for stitching via structure in the 25Gbps structure?
>>>>>>> 
>>>>>>>> "Lee " <leeritchey@xxxxxxxxxxxxx> 於 2014/9/7 上午3:17 寫道:
>>>>>>>> 
>>>>>>>> No stitching vias are needed.  Where did this idea get started?
>>>>>>>> 
>>>>>>>> -----Original Message----- From: Leeyuyun
>>>>>>>> Sent: Saturday, September 06, 2014 4:25 AM
>>>>>>>> To: si-list@xxxxxxxxxxxxx
>>>>>>>> Cc: lee.home.61@xxxxxxxxx
>>>>>>>> Subject: [SI-LIST] One stitching via or more vias is better for 25Gbps 
>>>>>>>> application???
>>>>>>>> 
>>>>>>>> Hi, all
>>>>>>>> I am doing 100Gbps (4x25Gps) product now
>>>>>>>> How do we put stitching vias when high speed signal change the layer?
>>>>>>>> This is a differential pair, which has P and N
>>>>>>>> Someone told me that one stitching vias is better because P and N 
>>>>>>>> share the common return
>>>>>>>> Someone told me that more stitching vias is better because loop 
>>>>>>>> inductance could be reduced more
>>>>>>>> I am confused which one is better?
>>>>>>>> Can anyone help me?
>>>>>>>> Thanks, 
>>>>>>>> Lee------------------------------------------------------------------
>>>>>>>> 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 forum  is accessible at:
>>>>>>>>             http://tech.groups.yahoo.com/group/si-list
>>>>>>>> 
>>>>>>>> List archives are viewable at:
>>>>>>>> http://www.freelists.org/archives/si-list
>>>>>>>> 
>>>>>>>> Old (prior to June 6, 2001) list archives are viewable at:
>>>>>>>> http://www.qsl.net/wb6tpu
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> -----
>>>>>>>> No virus found in this message.
>>>>>>>> Checked by AVG - www.avg.com
>>>>>>>> Version: 2014.0.4765 / Virus Database: 4015/8162 - Release Date: 
>>>>>>>> 09/05/14
>>>>>>> ------------------------------------------------------------------
>>>>>>> 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 forum  is accessible at:
>>>>>>>               http://tech.groups.yahoo.com/group/si-list
>>>>>>> 
>>>>>>> List archives are viewable at:
>>>>>>>       http://www.freelists.org/archives/si-list
>>>>>>> 
>>>>>>> Old (prior to June 6, 2001) list archives are viewable at:
>>>>>>>         http://www.qsl.net/wb6tpu
>>>>>> 
>>>>>> -- 
>>>>>> Steve Weir
>>>>>> IPBLOX, LLC
>>>>>> 1580 Grand Point Way
>>>>>> MS 34689
>>>>>> Reno, NV  89523-9998
>>>>>> www.ipblox.com
>>>>>> 
>>>>>> (775) 299-4236 Business
>>>>>> (866) 675-4630 Toll-free
>>>>>> (707) 780-1951 Fax
>>>>>> 
>>>>>> All contents Copyright (c)2013 IPBLOX, LLC.  All Rights Reserved.
>>>>>> This e-mail may contain confidential material.
>>>>>> If you are not the intended recipient, please destroy all records
>>>>>> and notify the sender.
>>>>>> 
>>>>>> ------------------------------------------------------------------
>>>>>> 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 forum  is accessible at:
>>>>>>              http://tech.groups.yahoo.com/group/si-list
>>>>>> 
>>>>>> List archives are viewable at:
>>>>>>       http://www.freelists.org/archives/si-list
>>>>>> 
>>>>>> Old (prior to June 6, 2001) list archives are viewable at:
>>>>>>       http://www.qsl.net/wb6tpu
>>>> 
>>>> 
>>>> -- 
>>>> Steve Weir
>>>> IPBLOX, LLC
>>>> 1580 Grand Point Way
>>>> MS 34689
>>>> Reno, NV  89523-9998
>>>> www.ipblox.com
>>>> 
>>>> (775) 299-4236 Business
>>>> (866) 675-4630 Toll-free
>>>> (707) 780-1951 Fax
>>>> 
>>>> All contents Copyright (c)2013 IPBLOX, LLC.  All Rights Reserved.
>>>> This e-mail may contain confidential material.
>>>> If you are not the intended recipient, please destroy all records
>>>> and notify the sender.
>>>> 
>>>> ------------------------------------------------------------------
>>>> 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 forum  is accessible at:
>>>>             http://tech.groups.yahoo.com/group/si-list
>>>> 
>>>> List archives are viewable at:     
>>>> http://www.freelists.org/archives/si-list
>>>> 
>>>> Old (prior to June 6, 2001) list archives are viewable at:
>>>> http://www.qsl.net/wb6tpu
>>> 
>>> 
>>> -- 
>>> Steve Weir
>>> IPBLOX, LLC 
>>> 1580 Grand Point Way
>>> MS 34689
>>> Reno, NV  89523-9998 
>>> www.ipblox.com
>>> 
>>> (775) 299-4236 Business
>>> (866) 675-4630 Toll-free
>>> (707) 780-1951 Fax
>>> 
>>> All contents Copyright (c)2013 IPBLOX, LLC.  All Rights Reserved.
>>> This e-mail may contain confidential material.
>>> If you are not the intended recipient, please destroy all records
>>> and notify the sender.
>> 
>> ------------------------------------------------------------------
>> 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 forum  is accessible at:
>>              http://tech.groups.yahoo.com/group/si-list
>> 
>> List archives are viewable at:     
>>       http://www.freelists.org/archives/si-list
>> 
>> 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 forum  is accessible at:
>               http://tech.groups.yahoo.com/group/si-list
> 
> List archives are viewable at:     
>        http://www.freelists.org/archives/si-list
> 
> 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 forum  is accessible at:
               http://tech.groups.yahoo.com/group/si-list

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
 
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu
  

Other related posts: