Walter,
Thanks for the thoughts. I have a little hard time with the "DC impedance"
terminology here.
If you look at the Wiki article:
https://en.wikipedia.org/wiki/Electrical_impedance
you will see that they say:
"In addition to resistance as seen in DC circuits, impedance in AC circuits
includes the..."
So I feel that at DC we should be talking about resistance, not impedance...
The other thing that still bothers me is your statement about Rail
connectivity. What does
"connected legally" mean?
If I were to summarize what you are saying, than the only difference between
PDN and signal
connectivity is that in PDN we require a low DC resistance, but in signal
connections we can
also consider a low AC impedance (at Nyquist) in addition to a low DC
resistance as a "good
connection".
Thanks,
Arpad
=========================================================================
From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] ;
On Behalf Of Walter Katz
Sent: Friday, December 13, 2019 9:08 AM
To: IBIS-ATM <ibis-macro@xxxxxxxxxxxxx>
Subject: [ibis-macro] What is a connection?
All,
I/O Connectivity
The IBIS specification defines the electrical IV and VT characteristics of an
I/O buffer. It is fundamentally based on the fact that there is an electrical
path that can be defined between a Driving Output buffer and one or more Input
(and other Output) buffers. It is expected that the transitions of the Driving
Output Buffer are reflecting in transitions at the Receiving Input buffers
between Vinl and Vinh. One way of describing these electrical connections
between the I/O buffers are with IBIS ISS (and Touchstone File) models. A
connections is a "good" connection if the transitions at the Driver generate
transitions at the Receiver that satisfy the receiver "rules" in the IBIS file.
A good connection could have a low DC impedance, or in some cases a low
impedance at a Nyquist frequency.
Rail Connectivity
If two nodes in a rail distribution system are "connected" legally then there
is a "low" DC impedance between the two nodes. Again, "low" is dependent on the
current t flow and proper operation of the I/O buffers.
Walter
Walter Katz
wkatz@xxxxxxxxxx<mailto:wkatz@xxxxxxxxxx>
Office 978.461-0449 x 133
Mobile 720.417-3762
[cid:image001.jpg@01D5B1B1.F7B54F00]