Tstonefile
The following section needs some discussion.
The default TX and RX node maps assume that the high-speed serial
channel
flows from left to right, and that analog model S parameters flow
from left
to right. Transmitter input ports are assumed to be the left hand (Near)
ports 1 and 3, while the pad ports are assumed to be the right hand
(Far)
ports 2 and 4. Receiver pad ports are assumed to be the left hand (Near)
ports 1 and 3, while the output ports are assumed to be the right
hand (Far)
ports 2 and 4.
Transmitter Default Port Map
Port Description
1 Input true port
2 Pad true port
3 Input complement port
4 Pad complement port
Receiver Default Port Map
1 Pad true port
2 Output true port
3 Pad complement port
4 Output complement port
Comments
1. In essence, the authors are saying that the network is
unidirectional, with no transmission coefficients in the reverse
direction. If this is the case, channel reflections will
terminate at the package/die pad boundary and not propagate into
the summer. (i.e. - real energy is lost in the simulation). As
long as the extraction of the analog circuit touchstone model was
performed at a sufficiently high impedance such that all
reflections from the hypothetical Tx or Rx summer have been
incorporated into the die-to-package path reflection coefficient
then there is not a problem. Effectively a large reflection
boundary is formed that meets the conditions outlined in the
Tstonefile discussion above. If not, if the die side ports are
terminated at a 50 ohm-ish characteristic impedance, then the
resulting reflected waveform will be incorrect.
2. Comment 1 necessarily requires clarification on Tx and Rx
termination, how it will be incorporated, and interaction with
touchstone file port impedance.
3. Port assignments seem to be oriented towards signal flow
direction, rather than universality at the die-package boundary.
I would argue for port numbering relative to the die pad and
hypothetical receiver/driver high impedance summer. In this way
the default node numbering is relative to the die and die pads,
just as package models should also be relative to the die pads and
bumps. Otherwise, mistakes in modeling will be made.
4. What about multiple channels for crosstalk modeling?
5. I'd argue for a similar construct for packages. Without a way to
encapsulate package models and concatenate them in the standard, a
significant part of the channel is missing.
6. Or ... if multiple channels are modeled with the Tstonefile
construct, then the model can be easily extended by the developer
to incorporate the complete cascade of the package and the die
analog models. To be consistent with IBIS, the package parasitics
in the IBIS package fields would be zeroed out.
Scott McMorrow
Teraspeed Consulting Group LLC
121 North River Drive
Narragansett, RI 02882
(401) 284-1827 Business
(401) 284-1840 Fax
Teraspeed® is the registered service mark of
Teraspeed Consulting Group LLC
On 10/6/2010 11:54 AM, Mirmak, Michael wrote:
>
> The enclosed BIRD 119, IBIS-AMI New Reserved Parameters, is
> distributed on behalf of Walter Katz, Mike Steinberger and Todd
> Westerhoff of SiSoft. It will be introduced at the next IBIS Open
> Forum meeting.
>
> A full list of BIRDs, their status and association with particular
> IBIS versions may be found at http://www.eda.org/ibis/birds/.
>
> Please note that the enclosed file is a "pointer" to a separate PDF
> file of the BIRD text, as the IBIS e-mail reflector does not permit
> redistribution of binary files. The enclosed file is identical to the
> ASCII file on the BIRD list at the URL above. The complete BIRD text
> is available at:
>
> http://www.eda.org/ibis/birds/bird119/bird119.pdf
>
> Michael Mirmak
>
> Intel Corp.
>
> Chair, IBIS Open Forum
>
>
> --
> This message has been scanned for viruses and
> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
> believed to be clean.
-- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------------------------------------------------------------- |For help or to subscribe/unsubscribe, e-mail majordomo@eda-stds.org |with the appropriate command message(s) in the body: | | help | subscribe ibis <optional e-mail address, if different> | subscribe ibis-users <optional e-mail address, if different> | unsubscribe ibis <optional e-mail address, if different> | unsubscribe ibis-users <optional e-mail address, if different> | |or e-mail a request to ibis-request@eda-stds.org. | |IBIS reflector archives exist under: | | http://www.eda-stds.org/pub/ibis/email_archive/ Recent | http://www.eda-stds.org/pub/ibis/users_archive/ Recent | http://www.eda-stds.org/pub/ibis/email/ E-mail since 1993Received on Wed Oct 6 16:31:42 2010
This archive was generated by hypermail 2.1.8 : Wed Oct 06 2010 - 16:33:05 PDT