Chris -- I see we're talking different languages. I'm talking about extracting parameters based on measurement, you're talking about extracting parameters based on field solution. While measurement-based extraction isn't always an option, in many cases it is. So, to make my assumptions more explicit: IF you are doing a measurement-based extraction (or an extraction based on 3-D results, such as PEEC, if you prefer), then in a number of applications (depending on edge rate and noise margins), using a TEM element (such as SPICE lossless TLines or HSPICE lossy W-Element) to model a trace propagating over munched-up package reference planes will be sufficiently accurate. Comparison of extracted model response vs. measurement will verify whether the accuracy is sufficient, or whether a more involved model is necessary. Our customers deal with these problems as well, so I agree it's a real-life problem and not a contrived one. And no, I wouldn't quibble about the results you have seen from field solver runs with various tools, particularly since I have no intention of replicating your effort... 8^) -- Steve ------------------------------------------- Steven D. Corey, Ph.D. Time Domain Analysis Systems, Inc. "The Interconnect Analysis Company." http://www.tdasystems.com email: steven.corey@xxxxxxxxxxxxxx phone: (503) 246-2272 fax: (503) 246-2282 ------------------------------------------- Chris Cheng wrote: > Noop, BTDT. Any TEM approximation will be way off. Especially when you > calculate crosstalk, impedance and prop delay (if the mesh is uniformly > defined). I am talking about 50%-200% error. If you don't believe me, try it > yourself. > > BTW, this is a well known problem. IBM engineers realise this way back when > they design the TCM modules for their mainframe. Albert Ruehli's PEEC method > works great on these problems. > > -----Original Message----- > From: Steve Corey [mailto:steven.corey@xxxxxxxxxxxxxx] > Sent: Thursday, September 18, 2003 11:32 PM > To: si-list@xxxxxxxxxxxxx > Subject: [SI-LIST] Re: TEM approximation > > > This is a good example of one of the basic assumptions underlying TEM > transmission line theory -- that the lines have to be uniform in the > direction of propagation. In other words, a 2D problem. > > A more interesting question from my perspective as a tool developer is > what is the "right answer" to which you refer. From an analytical > perspective, no tool will give the exact answer due to imprecision and > inherent model inaccuracies. From an engineering perspective, there are > many "right answers", since any answer that falls within an acceptable > error range of the exact answer is a right answer. So I can see where > in some applications a TEM approximation for the example below will be > fine and give a right answer, whereas in others it will not. It depends > how much accuracy you need to get a reliable design. > > Since our tools have both TEM and non-TEM capabilities, we recommend > that our customers use measurement validation to determine which > approach is necessary to get them the accuracy they need. > > -- Steve > > ------------------------------------------- > Steven D. Corey, Ph.D. > Time Domain Analysis Systems, Inc. > "The Interconnect Analysis Company." > http://www.tdasystems.com > > email: steven.corey@xxxxxxxxxxxxxx > phone: (503) 246-2272 > fax: (503) 246-2282 > ------------------------------------------- > > Chris Cheng wrote: > >>I can easily give you an example where TEM assumption will break down even >>with relatively slow edge rate. >> >>Take a simple uniform stripline. Start meshing the reference planes > > directly > >>below and above the trace. The size of the hole is dimensionally large >>compared with the trace itself but electrically small compared with the > > edge > >>rate so that quasi-static mode still holds. e.g. drill 50-100 mil holes >>uniformly spaced along a 5 mil stripline trace. No TEM mode tools or >>approximation will give you the right answer. The imagine current on the >>reference plane is forced to run with orthogonal components to the signal >>current, thus breaking the TEM mode. This is a very real life problem as > > it > >>exist in many signal traces in organic substrate packages. >> >>-----Original Message----- >>From: Dima Smolyansky [mailto:Dima.Smolyansky@xxxxxxxxxxxxxx] >>Sent: Wednesday, September 17, 2003 4:01 PM >>To: swamydp@xxxxxxxxx; si-list@xxxxxxxxxxxxx >>Subject: [SI-LIST] Re: TEM approximation >> >> >>Hello, Swami, >> >>Interestingly, we at TDA Systems determined that the TEM assumptions can > > be > >>used to surprisingly high risetimes if the parameters are accurately >>extracted and verified. Our lossy line examples verify this. >> >>I hope this helps. >> >>Thanks, >> >>-Dima >> >>----- Original Message ----- >>From: "Swamy Dhoss" <swamydp@xxxxxxxxx> >>To: <si-list@xxxxxxxxxxxxx> >>Sent: Sunday, September 07, 2003 4:28 PM >>Subject: [SI-LIST] TEM approximation >> >> >> >> >>>Hi >>> >>>Could anyone tell me about the concept of TEM >>>approximation for the propagation of an electrical >>>signal through a transmission line ? >>>Over what range of frequencies TEM approximation is >>>valid ? and what model can be used above this >>>frequency limit ? >>> >>>Regards >>>swamy >>> >>> >>>__________________________________ >>>Do you Yahoo!? >>>Yahoo! SiteBuilder - Free, easy-to-use web site design software >>>http://sitebuilder.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 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 >>> >>> >>> >> >> >>------------------------------------------------------------------ >>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 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 >> >>------------------------------------------------------------------ >>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 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 >> >> >> >> > > > -- ------------------------------------------------------------------ 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 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