Chris you are absolutely correct. That is why I said: "This return loss problem is always more acute for closely spaced "things" on low loss substrates, and will cause insertion loss uncertainty along with high frequency ringing." Chris Cheng wrote: > Scott, > If I recalled our off line conversation, I believe one of the problem is if > we purely just look at the S21 characteristics we will missed out the case > where the driving or receiving end is not ideally terminated at 50 ohm which > is likely the case at high gigabit design due to package and circuit > parasitics. If these caps and pads are placed close enough to the > receiver/driver the reflections can set up local resonance with each other > and result in ringing. > That is the reason why we agreed that placement of the series cap matters in > the channel even though that shouldn't be if the system is assumed to be > linear and perfectly terminated. > Relying purely on S21 without considering the driver/receiver condition is > not sufficient. > Regards, > Chris > > -----Original Message----- > From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On > Behalf Of Scott McMorrow > Sent: Friday, November 20, 2009 3:19 PM > To: Lee Ritchey > Cc: Yuriy Shlepnev; Steve Weir; Robert Haller; si-list@xxxxxxxxxxxxx > Subject: [SI-LIST] Re: FW: AC Blocking capacitor relative positions > andreferenceplanevoids > > To Lee and all, > I'd have to see your measurements and method to speak to your results > intelligently. It's been a while since I performed capacitor pad > simulations, optimizations and measurements. The results are similar to > what Yuriy shows. I currently have a need to do this for a product I am > designing for a client. This return loss problem is always more acute > for closely spaced "things" on low loss substrates, and will cause > insertion loss uncertainty along with high frequency ringing. I have > have an 0402 capacitor model built in CST (with plates) that has been > correlated to measurements, which I will use for a few modeling > experiments. Because the solver has been previously correlated for > this, and other similar problems, we can agree that it is a valid point > of comparison. I will report back on the results, and we will try to > include several examples in our next TRL test vehicle. > > Whether or not pad compensation with holes in the underlying planes > provides significant performance improvements depends upon the > dielectric thickness between the ground plane, the pads, and the > location of the lowest plate in the capacitor. In some cases, the > inductive loop between the capacitor plates and the PCB ground is > compensated by the excessive pad capacitance, up to reasonably high > frequencies. This may very well be the case that you happened on in > your test vehicle. In that case, your conclusions may quite valid. > > For others that may wish to try holes in the plane, be aware that > without a modeling environment and method that has been correlated to > measurements, you risk making your design worse. Placing holes in > ground planes is not for the weak of heart. It's best not to generalize > these sorts of design practices. > > Regards, > > Scott > > > > Lee Ritchey wrote: > >> Yuriy, >> >> I have measured data that trumps simulations every day. The reason I >> haven't published the data on this forum is it is contained in a large >> report that cannot be sent as an attachment. >> >> In virtually all cases, AC coupling capacitors are connected into nets >> using vias and surface mount pads. You are correct that the capacitance of >> these vias completely hides any effects of the mounting pads. The tests I >> have run are on nets with capacitors so mounted and the loss vs frequency >> measurements made up to 6 GHz. There is no detectable affect on this >> measurement with identical nets with and without the AC capacitors. The >> vias are 12 mil drill through a 100 mil thick PCB. There was no clearing >> of the planes under the capacitor mounting pads. >> >> >From this two conclusions can be drawn: >> >> 1. The disruptions associated with the vias and mounting pads is not >> significant, so why complicate the design job by insisting on extra artwork >> effort. >> >> 2. Since the disruptions are not significant, the physical location of the >> AC capacitor along the length of the path is not important. This had been >> pointed out by others on this forum such as Scott McMorrow. >> >> In general, I trust no simulations that are not validated with >> measurements, nor should anyone else. Until there is proof that the >> simulation model accurately represents the real circuit it should not be >> trusted. That is why I build so many test PCBs. I need to prove to my >> clients that the advice I am giving them is valid. After all, some of them >> bet in excess of $100 Million on this advice and it needs to be right. >> >> Anyone else who gives advice based solely on simulations needs to add that >> caveat to their advice. >> >> Anyone who gives advice simply because it is published in an applications >> note is in the same position. We all know how erroneous most application >> notes are. >> >> Lee Ritchey >> >> >> >> >>> [Original Message] >>> From: Yuriy Shlepnev <shlepnev@xxxxxxxxxxxxx> >>> To: Lee Ritchey <leeritchey@xxxxxxxxxxxxx>; Steve Weir <weirsi@xxxxxxxxxx> >>> Cc: Robert Haller <rhaller@xxxxxxxxxxxxx>; <si-list@xxxxxxxxxxxxx> >>> Date: 11/20/2009 12:40:31 PM >>> Subject: RE: [SI-LIST] Re: FW: AC Blocking capacitor relative positions >>> >>> >> andreferenceplanevoids >> >> >>> Lee, >>> >>> I would like to investigate your board with the AC caps numerically and >>> publish the results on SI list. Please send me brd or mcm file and point >>> >>> >> at >> >> >>> the nets to simulate. >>> >>> In general, the effect of the voids under AC caps may be not important, if >>> the pads are small or there are larger discontinuities in the channel. >>> >>> >> Vias >> >> >>> on both sides from the AC cap pads may be such larger discontinuity. Any >>> optimization of the pads may not provide any visible improvement because >>> >>> >> of >> >> >>> the vias. Smaller capacitor footprint and longer line segments may also >>> >>> >> lead >> >> >>> to lower reflections and make the optimization useless - see for instance >>> 0402 case in >>> http://www.simberian.com/AppNotes/AC_CouplingCapacitors_2008_04.pdf (slide >>> 15 - the reflection in non-optimized case is about -20 dB that may be >>> >>> >> hardly >> >> >>> noticeable on TDR). However, the larger footprint like 0603 on slide 16 >>> >>> >> with >> >> >>> the original reflection above -10 dB should be clearly visible on TDR. The >>> reflection in this case will depend on the position of the caps in the >>> channel. Again, the larger discontinuities like vias may obfuscate the >>> effect. There is always some advantage in the optimization, but the >>> >>> >> overall >> >> >>> effect may depend on a particular net. >>> >>> Best regards, >>> Yuriy Shlepnev >>> www.simberian.com >>> >>> >>> -----Original Message----- >>> From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] >>> >>> >> On >> >> >>> Behalf Of Lee Ritchey >>> Sent: Friday, November 20, 2009 9:28 AM >>> To: Steve Weir >>> Cc: Robert Haller; si-list@xxxxxxxxxxxxx >>> Subject: [SI-LIST] Re: FW: AC Blocking capacitor relative positions >>> andreferenceplanevoids >>> >>> Steve, >>> >>> I thought you said you had seen this phenomena and would share your >>> >>> >> results >> >> >>> with us. Isn't that what you stated in one of your replies? >>> >>> I'm willing to share my results with anyone who wants to send me a >>> request. I may regret this offer if I get too many requests! >>> >>> Lee Ritchey >>> >>> >>> >>> >>>> [Original Message] >>>> From: steve weir <weirsi@xxxxxxxxxx> >>>> To: <leeritchey@xxxxxxxxxxxxx> >>>> Cc: Robert Haller <rhaller@xxxxxxxxxxxxx>; si-list@xxxxxxxxxxxxx >>>> >>>> >>> <si-list@xxxxxxxxxxxxx> >>> >>> >>>> Date: 11/19/2009 5:40:45 PM >>>> Subject: [SI-LIST] Re: FW: AC Blocking capacitor relative positions >>>> >>>> >>> andreferenceplanevoids >>> >>> >>>> Lee, sure we can do that. I think it would be educational all the way >>>> around. I'd like to see the PCB files first. Are they available in >>>> Allegro? >>>> >>>> Best Regards, >>>> >>>> >>>> Steve. >>>> Lee Ritchey wrote: >>>> >>>> >>>>> I'd like to see that as my measurements don't show any significant >>>>> >>>>> >>> impact. >>> >>> >>>>> I've got test PCBs that can used to check. >>>>> >>>>> >>>>> >>>>> >>>>> >>>>>> [Original Message] >>>>>> From: steve weir <weirsi@xxxxxxxxxx> >>>>>> To: Lee Ritchey <leeritchey@xxxxxxxxxxxxx> >>>>>> Cc: Robert Haller <rhaller@xxxxxxxxxxxxx>; si-list@xxxxxxxxxxxxx >>>>>> >>>>>> >>>>>> >>>>> <si-list@xxxxxxxxxxxxx> >>>>> >>>>> >>>>> >>>>>> Date: 11/18/2009 7:49:18 PM >>>>>> Subject: [SI-LIST] Re: FW: AC Blocking capacitor relative positions >>>>>> >>>>>> >>>>>> >>>>> andreferenceplane voids >>>>> >>>>> >>>>> >>>>>> Lee, the difference is readily seen in VNA measurements. Depending >>>>>> >>>>>> >> on >> >> >>>>>> the signal spectra the eye is visibly impacted. >>>>>> >>>>>> Best Regards, >>>>>> >>>>>> >>>>>> Steve. >>>>>> Lee Ritchey wrote: >>>>>> >>>>>> >>>>>> >>>>>>> This could be one of those cases where the effect was visible, but >>>>>>> >>>>>>> >> not >> >> >>>>>>> significant. >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> [Original Message] >>>>>>>> From: Haller, Robert <rhaller@xxxxxxxxxxxxx> >>>>>>>> To: si-list@xxxxxxxxxxxxx <si-list@xxxxxxxxxxxxx> >>>>>>>> Date: 11/18/2009 1:34:29 PM >>>>>>>> Subject: [SI-LIST] FW: AC Blocking capacitor relative positions >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> andreference plane voids >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> JF, >>>>>>>> Depending on your specific high speed link requirements, voids >>>>>>>> >>>>>>>> >> under >> >> >>>>>>>> >>>>> the >>>>> >>>>> >>>>> >>>>>>>> >>>>>>>> >>>>>>> capacitors are an easy way to reduce the impedance discontinuity. We >>>>>>> >>>>>>> >>>>>>> >>>>> model >>>>> >>>>> >>>>> >>>>>>> this with a 3D Full Wave field solver. We also built up some >>>>>>> >>>>>>> >>> structures >>> >>> >>>>>>> >>>>> in >>>>> >>>>> >>>>> >>>>>>> the lab and measured them (both in time and frequency Domains). >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> We were pleasantly surprised how much the reliefs helped and how >>>>>>>> >>>>>>>> >> well >> >> >>>>>>>> >>>>> the >>>>> >>>>> >>>>> >>>>>>>> >>>>>>>> >>>>>>> correlation to tools were. I wouldn't worry about the difference >>>>>>> >>>>>>> >>>>>>> >>>>> between a >>>>> >>>>> >>>>> >>>>>>> two cutouts versus a single void. If you can make the capacitor look >>>>>>> >>>>>>> >>>>>>> >>>>> like a >>>>> >>>>> >>>>> >>>>>>> 100Ohm transmission line (with no impedance discontinuity) then the >>>>>>> placement isn't critical, but in practice it depends. I have seen >>>>>>> >>>>>>> >>> some >>> >>> >>>>>>> vendors specify lengths to avoid, because if the package length plus >>>>>>> distance to the cap (and the associated discontinuity) are a >>>>>>> >>>>>>> >> multiple >> >> >>> of >>> >>> >>>>>>> the UI (unit interval of the LINK) undesirable resonances can occur. >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> Regards >>>>>>>> Bob Haller >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> -----Original Message----- >>>>>>>> From: si-list-bounce@xxxxxxxxxxxxx >>>>>>>> >>>>>>>> >>>>>>>> >>>>> [mailto:si-list-bounce@xxxxxxxxxxxxx] >>>>> >>>>> >>>>> >>>>>>>> >>>>>>>> >>>>>>> On Behalf Of jean-francois hasson >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> Sent: Wednesday, November 18, 2009 4:04 PM >>>>>>>> To: si-list@xxxxxxxxxxxxx >>>>>>>> Subject: [SI-LIST] AC Blocking capacitor relative positions and >>>>>>>> >>>>>>>> >>>>>>>> >>>>> reference >>>>> >>>>> >>>>> >>>>>>>> >>>>>>>> >>>>>>> plane voids >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> Hi, >>>>>>>> >>>>>>>> When looking at AC coupling capacitors on high speed serial links >>>>>>>> >>>>>>>> >> it >> >> >>>>>>>> >>>>> is >>>>> >>>>> >>>>> >>>>>>>> often recommended to void the reference plane under these >>>>>>>> >>>>>>>> >> capacitors >> >> >>>>>>>> >>>>> to >>>>> >>>>> >>>>> >>>>>>>> minimize capacitance variations. Most of the times, in the designs >>>>>>>> >>>>>>>> >> I >> >> >>>>>>>> have seen, the void is exactly underneath each AC blocking >>>>>>>> >>>>>>>> >> capacitor >> >> >>>>>>>> including landing pads. If ever the capacitors were quite close, >>>>>>>> >>>>>>>> >>> would >>> >>> >>>>>>>> there be any reason why a single void below both capacitors would >>>>>>>> >>>>>>>> >> be >> >> >>>>>>>> >>>>> an >>>>> >>>>> >>>>> >>>>>>>> issue ? I have also noticed that most of the times AC blocking >>>>>>>> capacitors are not as close as possible to a transmitter or >>>>>>>> >>>>>>>> >> receiver >> >> >>>>>>>> where I thought it was beneficial to do so to remove the number of >>>>>>>> impedance disruptions ? Could anyone provide me with some >>>>>>>> >>>>>>>> >>> explanation ? >>> >>> >>>>>>>> Best regards, >>>>>>>> >>>>>>>> JF >>>>>>>> >>>>>>>> ------------------------------------------------------------------ >>>>>>>> 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 >>>>>>>> >>>>>>>> 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: >>>>>>>> //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 >>>>>>>> >>>>>>>> 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: >>>>>>> //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 >>>>>>> >>>>>>> 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: >>>>>> //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 >>>>>> >>>>>> 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: >>>> //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 >>>> >>>> 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: >>> //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 >>> >>> 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: >> //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 >> >> Old (prior to June 6, 2001) list archives are viewable at: >> http://www.qsl.net/wb6tpu >> >> >> >> >> > > -- > 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(r) is the registered service mark of > Teraspeed Consulting Group LLC > > > > ------------------------------------------------------------------ > 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 > > Old (prior to June 6, 2001) list archives are viewable at: > http://www.qsl.net/wb6tpu > > > > This email and any attachments thereto may contain private, confidential, and > privileged material for the sole use of the intended recipient. Any review, > copying, or distribution of this email (or any attachments) by others is > strictly prohibited. If you are not the intended recipient, please contact > the sender immediately and permanently delete the original and any copies of > this email and any attachments thereto. > ------------------------------------------------------------------ > 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 > > Old (prior to June 6, 2001) list archives are viewable at: > http://www.qsl.net/wb6tpu > > > > -- 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 ------------------------------------------------------------------ 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 Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu