RE: [10GBASE-CX4] S-param characterization of the cable

["Nadolny, Jim" <JNadolny@fciconnect.com>]

John,

Good catch and thanks.  

We use the SOLT cal and the mapping is indeed 1-3-2-4 as you describe, not
the 1-2-4-3 as I indicated.  I knew it was non-standard but had ports 2 and
3 swapped.

regards,
Jim N

-----Original Message-----
From: DAmbrosia, John F [mailto:john.dambrosia@tycoelectronics.com]
Sent: Wednesday, June 18, 2003 11:38 AM
To: 'Nadolny, Jim'; Oleynick, Gary; 'Zhi Wong'; 'DOVE,DANIEL J
(HP-Roseville,ex1)'; 'Liu, Howard'; '10GBASE-CX4 (E-mail)'
Subject: RE: [10GBASE-CX4] S-param characterization of the cable


Jim,
I just looked into this issue, and per Agilent, the port assignment is
dependent on the calibration scheme used.

the PLTS uses the following balanced port
configuration when using the SOLT calibration 
> Port 1 - In +
> Port 3 - In -
> Port 2 - Out +
> Port 4 - Out -

When using the TRL calibration the port configuration is as follows

>Port 1 - In +
> Port 2 - In -
> Port 3 - Out +
> Port 4 - Out -


John D'Ambrosia
Manager, Semiconductor Relations
Tyco Electronics Corporation

Tel. 717.986.5692
Mobile 717.979.9679

email - john.dambrosia@tycoelectronics.com

-----Original Message-----
From: Nadolny, Jim [mailto:JNadolny@fciconnect.com]
Sent: Wednesday, June 18, 2003 11:33 AM
To: Oleynick, Gary; 'Zhi Wong'; 'DOVE,DANIEL J (HP-Roseville,ex1)';
'Liu, Howard'; '10GBASE-CX4 (E-mail)'
Subject: RE: [10GBASE-CX4] S-param characterization of the cable



Gary,

Using the 4 port VNA from Agilent we have ports 1 and 2 as inputs and ports
4 and 3 as outputs.  The through measurements are 1 to 4 and 2 to 3.  The
mode conversion equations built into the Agilent software assume this
configuration when it derives the mixed mode S-parameters.  Using this
instrument it is most efficient to simply follow their convention.

We can re-map the ports so that a more academic (and logical) convention is
followed.  In data that I provide I try to define the port convention so
that the user can re-map if they want to.  This requires the user to
actually read the comments in the Touchstone file to understand the
convention.  I have seen other mappings as well (1 and 2 are inputs and 3
and 4 are outputs, S13 and S24 are the through connections) so until a
standard is applied, I will include a port description in the comment line.

Dan - For 10 GBE work it would make sense to mandate a standard port
mapping.  The 1-2-4-3 mapping is easiest for me but the 1-3-2-4 or 1-2-3-4
are more logical.

regards,
Jim N

-----Original Message-----
From: Oleynick, Gary 
Sent: Wednesday, June 18, 2003 11:20 AM
To: 'Zhi Wong'; DOVE,DANIEL J (HP-Roseville,ex1); Liu, Howard; Oleynick,
Gary; 10GBASE-CX4 (E-mail); Nadolny, Jim
Subject: RE: [10GBASE-CX4] S-param characterization of the cable


Jim,

How should we respond to this request?

Gary 

-----Original Message-----
From: Zhi Wong [mailto:zwong@altera.com]
Sent: Tuesday, June 17, 2003 8:43 PM
To: DOVE,DANIEL J (HP-Roseville,ex1); Liu, Howard; Oleynick, Gary;
10GBASE-CX4 (E-mail)
Subject: RE: [10GBASE-CX4] S-param characterization of the cable


Dan/Howard/Gary,

I would suggest taking 4-port S-Parameter measurements.
(Ports 1 and 3 on the left of the network)
(Ports 2 and 4 on the right of the network)

This would be consistent with some OIF standards like CEI and XFP.

Zhi
Altera Corporation
San Jose, CA
USA

-----Original Message-----
From: DOVE,DANIEL J (HP-Roseville,ex1) [mailto:dan.dove@hp.com] 
Sent: Tuesday, June 17, 2003 4:45 PM
To: 'Liu, Howard'; 'Oleynick, Gary'; 10GBASE-CX4 (E-mail)
Subject: RE: [10GBASE-CX4] S-param characterization of the cable



Hello,

For those who are observing this conversation, the data Gary sent
exceeded the 100K maximum size limit for the reflector. As a result,
only those directly addressed received the file.

Regards,

Dan Dove
Chair, 802.3ak


> -----Original Message-----
> From: Liu, Howard [mailto:howard.liu@ti.com]
> Sent: Tuesday, June 17, 2003 4:26 PM
> To: 'Oleynick, Gary'; 10GBASE-CX4 (E-mail)
> Subject: RE: [10GBASE-CX4] S-param characterization of the cable
> 
> 
> 
> Gary,
> This is a complete measurement, the format is good, I can
> convert single
> ended data to differential data. Do you have the rest of the 
> measurements.
> This is for the test fixture only.
> 
> Thanks!
> 
> howard liu
> 
> 
> -----Original Message-----
> From: Oleynick, Gary [mailto:GOleynick@fciconnect.com]
> Sent: Tuesday, June 17, 2003 6:16 PM
> To: Liu, Howard; Oleynick, Gary; 10GBASE-CX4 (E-mail)
> Subject: RE: [10GBASE-CX4] S-param characterization of the cable
> 
> 
> Howard,
> 
> Does this help?
> 
> Gary
> 
> 
> 
> -----Original Message-----
> From: Liu, Howard [mailto:howard.liu@ti.com]
> Sent: Tuesday, June 17, 2003 4:08 PM
> To: 'Oleynick, Gary'; 10GBASE-CX4 (E-mail)
> Subject: RE: [10GBASE-CX4] S-param characterization of the cable
> 
> 
> Gary,
> Thanks again, I looked into the files, looks like that it only has the

> amplitude measurements in dBs, I need its phase measurements too if 
> you have them.
> 
> Thanks!
> 
> howard
> 
> 
> -----Original Message-----
> From: Oleynick, Gary [mailto:GOleynick@fciconnect.com]
> Sent: Tuesday, June 17, 2003 5:02 PM
> To: Liu, Howard; 'DOVE,DANIEL J (HP-Roseville,ex1)'; 10GBASE-CX4
> (E-mail)
> Cc: Oleynick, Gary
> Subject: RE: [10GBASE-CX4] S-param characterization of the cable
> 
> 
> Howard,
> 
> Attached is FCI's S-parameter data on the 10 and 15 meter
> cable assembly
> lengths.
> 
> Regards,
> 
> Gary
> 
> 
> 
> -----Original Message-----
> From: Liu, Howard [mailto:howard.liu@ti.com]
> Sent: Tuesday, June 17, 2003 2:31 PM
> To: 'DOVE,DANIEL J (HP-Roseville,ex1)'; 10GBASE-CX4 (E-mail)
> Subject: [10GBASE-CX4] S-param characterization of the cable
> 
> 
> 
> 
> Hi,
> Do we have an exist measurement of the channel (cable) from 
> TP1 to TP4 with
> FEXT and NEXT included? It can be in any format. 
> 
> Thanks!
> 
> 
> Howard Liu
> 
> Texas Intruments
> 214 480 7678
>