[ibis-macro] Re: another silly IBIS-AMI question about deriving the impulse response of a channel

  • From: Mike Steinberger <msteinb@xxxxxxxxxx>
  • To: scott@xxxxxxxxxxxxx
  • Date: Thu, 27 May 2010 17:45:09 -0500

Scott-

What leads to believe that the network has to stop at the Rx pad? If there are significant circuit elements between the Rx pad and the Rx termination, what leads you to believe that they would be excluded from the network analysis?

Please be aware that over a year ago, we proposed to IBIS ATM using on-die S parameters to solve a number of problems, including proper modeling of the on-die circuit elements. We've worked with several companies who use on-die S parameters because they feel it's a better way to model the analog behavior of the pin electronics. To date, such models must be considered to be proprietary and non-compliant with IBIS because there has been no action on the on-die S parameter proposal.

I also wonder what you mean by "large errors". What is your numerical estimate of the magnitude of such errors? How did you perform that estimate? What data did you base your estimate on?

Mike S.

On 05/27/2010 05:31 PM, Scott McMorrow wrote:
Walter

you would be incorrect.
"Lumped approximations between TxTerm/TxPad and RxTerm/RxPad do not hold"

A simulation based upon your assumption would have large errors.

Scott




On 5/27/2010 6:31 PM, Walter Katz wrote:

Scot,

I think a better picture is:


TxTerm --- TxDie --- TxBump -- TxPkg -- channel -- RxPkg -- RxBump -- RxDie --- RxTerm
  | |
  |   |
TxPad RxPad

Walter

Walter Katz

303.449-2308

Mobile 720.333-1107

wkatz@xxxxxxxxxx

www.sisoft.com

-----Original Message-----
*From:* Scott McMorrow [mailto:scott@xxxxxxxxxxxxx]
*Sent:* Thursday, May 27, 2010 6:25 PM
*To:* Walter Katz
*Cc:* IBIS-ATM
*Subject:* Re: [ibis-macro] another silly IBIS-AMI question about deriving the impulse response of a channel

Walter,

You are misinterpreting my picture. I was showing the network of just a Tx or Rx. When I say "network" I truly mean a circuit network. Let me create a better picture of the entire channel.

TxTerm --- TxDie --- TxBump -- TxPkg -- channel -- RxPkg -- RxBump -- RxDie --- RxTerm
             |                                                          |
             |                                                          |
TxPad RxPad


For a 50 ohm single-ended channel

TxTerm = 50 ohm
RxTerm = 50 ohm
TxPad = some high impedance
RxPad = some high impedance
Lumped approximations between TxTerm/TxPad and RxTerm/RxPad do not hold

What I was showing was just one part of the total channel.

(50 ohm)RxBump --- Network --- RxTerm (50 ohm)
                      |
                      |
                    RxPad
                   (10k ohm)


--
Scott McMorrow
Teraspeed Consulting Group LLC
121 North River Drive
Narragansett, RI 02882
(401) 284-1827 Business
(401) 284-1840 Fax

http://www.teraspeed.com

Teraspeed® is the registered service mark of
Teraspeed Consulting Group LLC

Other related posts: