[SI-LIST] : Partial inductance....

About this list Date view Thread view Subject view Author view

From: rich.evans@st.com
Date: Tue Mar 20 2001 - 07:51:36 PST


     Michael, thank you for the very clear discussion.
     
     
     After many hours (years?) of messing with inductances, I
     must agree completely with Michael. Partial inductance
     can be useful, but it's very dangerous.
     
     best regards to all
     
     
     Rich
     ----------------------------------------------------------------------
     
     
     
     Sainath, I used to believe as you do, that partial inductances are
     useful to obtain some first-cut answers. Over the years, I've changed
     my mind. I believe that the potential for misuse from partial
     inductances outweighs their benefits, and I'm now doing all my signal
     integrity modeling with loop inductances. I'm much happier. :-)
     Here are some of the problems I see with partial inductances: 1.)
     They are arbitrary; as Brian Young points out in his wonderful new
     book, you can add any constant you want to the partial inductance
     matrix without changing the physical result. Different techniques for
     calculating partial inductance give different answers --- witness the
     discussion we've just had on this point. 2.) When you use partial
     inductances in SPICE simulations, they give you things that look like
     "ground bounce": voltage differences across large sections of your
     circuit, where it is impossible to make a unique physical measurement
     of voltage because of linked flux. Brian Young again points out that
     ground bounce is not unique; it depends on your definition of partial
     inductance. You can be mislead by how chip ground is bouncing with
     respect to module ground in your simulation --- it looks like
     something real, but it's not. When you use loop inductances, and use
     SPICE node 0 to represent local reference everywhere, you can't be
     mislead; there's no node voltage in your simulation that looks like
     ground bounce. 3.) If you use partial inductances in your SPICE
     simulations, you have to make sure that all the current in your
     simulation moves from one side of your circuit to the other only
     through the partial inductances. If you have node 0 on both sides,
     for example, you've violated the assumptions under which partial
     inductance is valid. And it can be very hard to avoid node 0
     sometimes, and it appears that having large sections of your circuit
     isolated from node 0 makes convergence more difficult. 4.) Partial
     inductances are completely invalid without mutual inductances, but
     there's a great tendency to ignore them as a "first-pass engineering
     assumption". This is natural; all of engineering is about ignoring
     things. :-) But it just doesn't work with partial inductances. At
     best, you're making assumptions about where the return path is (and
     different ways of calculating partial inductances make different
     assumptions); at worst, you miss the entire point of the exercise.
     Without partial mutual inductances, there's no reason to put power and
     ground planes close to each other. Basically, my feeling now is that
     partial inductances are a wonderful tool for calculating inductance in
     the standard signal integrity situation where the full loop is not
     completely known (package without chip or board, for example). But I
     think now they should remain a computational tool, and that the models
     that are eventually generated should be based on loop inductances.
     I'm working on a paper explaining these points in more detail and
     talking about how we've been using loop inductance rather than partial
     inductance for package modeling here at Compaq. I hope to present the
     paper at EPEP'01 here in Massachusetts. I would appreciate any
     comments people might have.
     
     --
     Michael Tsuk
     Compaq AlphaServer Product Development
     (508) 467-4621 -----Original Message-----
     From: Sainath Nimmagadda [mailto:sainath@lsil.co

**** To unsubscribe from si-list or si-list-digest: send e-mail to
majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Jun 21 2001 - 10:11:17 PDT