All,
If you made the second case pins 456, then pin 4 would have been an
aggressor in two models and a victim in no model. This is the case that
Arpad wanted to document
Pins 2 and six would be each in one model as an aggressor. Pins 3 and 5
would be each be in one model as an aggressor.
Walter
Walter Katz
<mailto:wkatz@xxxxxxxxxx> wkatz@xxxxxxxxxx
978.461-0449 x 133
Mobile 303.335-6156
From: ibis-interconn-bounce@xxxxxxxxxxxxx
[mailto:ibis-interconn-bounce@xxxxxxxxxxxxx] On Behalf Of Mike LaBonte
Sent: Wednesday, January 10, 2018 4:24 PM
To: ibis-interconn@xxxxxxxxxxxxx
Subject: [ibis-interconn] Aggressor_Only graphic version 4 for discussion
The attached version 4 Aggressor_Only graphic has the "aggressor"
checkboxes eliminated and the label for the "victim" checkboxes enhanced
to form a sort of legend, as agreed in the Tuesday IBIS ATM meeting. The
following highlighted update is proposed for the BIRD 198.x Aggressor_Only
description, to introduce the figure:
Multi-line models may describe only a subset of a coupled structure (e.g.,
a 64-line bus may be described by a four-line model). As a result, while
the interconnects at the edges of the model may induce crosstalk onto
other interconnects nearby, being on the edge of the model, they may not
themselves experience the full crosstalk impact that the corresponding
interconnect experiences in the real, full structure. Figure xx shows
examples of interconnect models having full coupling for some pins and
partial coupling for other pins of an example part package, and the
corresponding Aggressor_Only entries. The Aggressor_Only column entry is
allowed on all terminal locations for I/O terminals to indicate such
incomplete coupling. Terminals that include the Aggressor_Only entry may
not be suitable to be simulated as victims, as they do not experience the
full coupling present in the real physical structure. If an I/O terminal
is not identified as Aggressor_Only, then the interconnect to that I/O
terminal includes coupling to all interconnections deemed necessary for
coupled signal analysis. If a particular terminal is identified as
Aggressor_Only, then the entire path of the associated pin_name is to be
considered Aggressor_Only.
The graphic would appear after that paragraph and before the header
"Connecting Pins, Pads and Buffer Terminals". I will incorporate the
figure and text into a draft 16 of BIRD189.5, once draft 15 has been
established as the next baseline. We may have other related figure
proposals.
Mike
