Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3_MAINT] ICN ad hoc meeting minutes for Tues Feb 21



Hello Tom

 

I expect that due care will be taken to limit the ICN related changes to the CXP connector.  I expect requirements placed on the QSFP+ connector will not modified.

 

Regards,

 

John

 

From: Tom Palkert [mailto:tpalkert@xxxxxxxx]
Sent: Tuesday, February 21, 2012 2:35 PM
To: STDS-802-3-MAINT@xxxxxxxxxxxxxxxxx
Subject: [802.3_MAINT] ICN ad hoc meeting minutes for Tues Feb 21

 

ICN ad hoc conf call Feb 21 2012:

 

Attendees: Tom Palkert (Molex), Mike Dudek (Qlogic), Piers Dawe (IPtronics), Nathan Tracy (TE connectivity), Adam Healey (LSI), Kent Lusted (Intel), John Petrilla (Avago), Brian Misek (Avago),  Pete Anslow (Ciena)

 

Summary: The group was generally supportive of the proposed ICN changes and agreed that the impact to the jitter specifications would be minimal and did not warrant any changes to the jitter specs.  The proposed Qsq change from 45 to 40 may be too aggressive and a value of 41-42 may be more appropriate.

 

 

Agenda:

1)Review Mike Dudek revised spreadsheet on jitter impacts of proposed ICN change

2) Review Piers presentation from Quebec 2009 explaining the need for Qsq.

3) Review Piers spreadsheet showing the impacts of a change in Qsq from 45 to 40.

 

1)      Mike showed that the increased jitter measured at TP1a caused by the propose change in ICN would be minimal ( increase of 1% after changing Y2 amplitude to 350mv per Piers spreadsheet).

2)      Piers reviewed his presentation that formed the basis for the original Qsq spec. There was a discussion about the appropriate bandwidth that was assumed to represent the host Receiver function. (Clause 86A) There was no consensus on the correct BW (7.5GHz vs 12GHz). This will be discussed at our next conf. call.

3)      Piers reviewed his spreadsheet showing the effects of reducing the Qsq value from 45 to 40.

Comment from Mike on  Column B: He believes that the integrated crosstalk values are too large (they assume a 1200mv amplitude)

Piers used  350mv amplitude for Y2. Mike agreed that this was a correct value and will update his spreadsheet accordingly.

 

 

Next meeting action items and agenda:

Piers to review his spreadsheet to determine if he is too pessimistic in some of his assumptions on slew rate vs worst case eye.

Tom: Present measured data showing eye diagrams with and without crosstalk present.

Tom: Calculate proposed ICN changes with both 7.5GHz filter  and 12GHz filter.

Pete: Lead discussion on proposed changes to 86A if filter bandwidth is changed.

Next meeting is Tues Feb. 28 at 11am Pacific. Call in info (same as last time):

Call in info:

1-800-503-2784

Code 7185106

Time: 11am Pacific Tues Feb 28.

 

 

Thanks for your participation

Tom Palkert

ICN ad hoc chair