Re: bird 28 comments

Let me tell you what I think I would do.

For any group of lines that couple you need an "ACTIVE" line, that is, the one that
is going to be stimulated. This is usually the line in the middle of a group since outside lines
would, per se, couple to more lines surrounding them if they were ACTIVE. Now you build your simulation
net list. I would do something like:

P1 ===========================
A ========================
P2 ====================

  |-------------------|---|--|

(A is active net and P1 and P2 are passive nets. Pretend they are at 45% angle as in Stephens drawing)

Net A would couple to P2 and P1 using a 3 config coupling for the length of P2. Net A would couple to net
P1 with a 2 config coupling from the end of P2 to the end of A. Net P1 would have an additional, "self only"
coupling for the rest of its length. Ideally, IBIS would supply this data, baring that, perhaps things could
be hacked. Of course, if we know how people are going to hack things we should just give the right data
to begin with.

jon
Received on Wed Jul 19 19:19:41 1995