[8023-POEP] Unapproved minutes from 350uH ad hoc technical sub-call held 6/10/08

[George Zimmerman <gzimmerman@xxxxxxxxxxxxxx>]

Following are the notes from the 350uH sub call held on 6/10/08 to reconcile technical approaches to analysis and laboratory proof, offline from the main group  That goal appears to have been achieved. These minutes have been reviewed by the attendees via email and have attained email consensus, but a formal approval was not taken.  They are provided FYI for other members of the 350uH ad hoc who were unable to make the call.

-george zimmerman

Date: 6/10/08

Attendees:

George Zimmerman, George Eisler, Chris Pagnanelli – Solarflare

Mike Grimwood – Broadcom

Yair Darshan – Microsemi

(Fred Schindler joined at the end)

 

At 8:08 AM PDT, the attendees were asked if they had read and understood the IEEE Patent policy, they confirmed.  There were no disclosures.

 

Yair opened with a summary of the understanding reached by the email discussion:

-          He had tried to use the tools from midspan tools to check how sensitive the channel is to changing the OCL.

-          These indicated that the System Channel Model changes are small (~1dB or less).  However, based on our discussion his thinking has moved to checking the transient response, rather than the steady state response that the System Channel Model reflects.

-          His intent is to look at different OCLs to determine the minimum OCL, with his key question as to whether we could go to the 70% droop, something intermediate, or do we need to stay at 80% droop.

 

George Zimmerman clarified that the UNH tests were performed on 40uH OCL with long-term droop compensation, and that we did not look at intermediate droop values between test #3 (droop equiv to 170uH) and 350uH.

 

There was some discussion where Yair focused on DUT #6 and whether it might be an invalid data point.  He suggested that he might ask UNH to pursue further testing of DUT#6, at least to determine it’s sensitivity vs. OCL.

 

Mike Grimwood  clarified that the two questions we were considering were:

1. Can a droop specification less than 87% be used?
2. Can we exclude a device from testing.

 

There was further discussion that we would not want to ignore a device that failed, but would likely pursue further testing.

 

Yair volunteered to go back to UNH and ask them to test sensitivity vs. OCL for DUT #6, and, based on those results might run further tests at an intermediate OCL.

 

Yair polled the group as to whether they would consider changing the specification to be equivalent to a lower inductance if tests show that dut#6 passes with an inductance between 170uH and 350uH .  3 members responded affirmatively, with discussion suggesting that they would consider it, but more tests would be required to actually confirm such a decision.

 

Yair asked whether the issue of magnetics saturation by the test waveform had been resolved.  Chris Pagnanelli offered that the vendor who had responded to him had tested both in the lab and in simulation and found no problem.  He will try to get a contribution from the vendor.

 

Yair suggested that George ask Mike  & David Law for a 2 hour block before the meeting in New Hampshire to close this alternative spec discussion.

 

 

We then reviewed the notes and the following action items:

1. Yair to ask for further tests at UNH-IOL to check the sensitivity of DUT #6 to OCL.
2. Chris to ask magnetics vendors for a contribution (or other documented closure) of the saturation concern.
3. George to ask David/Mike for time at the meeting.

 

Fred joined the call during the notes review and asked a question regarding the length of the killer packet response used in the tests.  It is much longer than the DDJ test pattern.  George deferred to Dan Dove and the two agreed to take the subject offline.

 

Call was adjourned at 8:58 AM PDT.

 

 

 

George A. Zimmerman

Founder & CTO, PHY Technologies

Solarflare Communications

gzimmerman@solarflare.com

Tel:  949-581-6830 x2500

Cell: 310-920-3860