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

Re: [802.3_4PPOE] Analysis of E2EP2PRUNB with Dual Signature PDs

Yair -

Page 7 (Figure 3c) is a plausible connection, but shouldn’t it be treated the same as Figure 1 by the PSE (if the PSE has the positive path pairs tied together, like most do)? Most PSEs won’t measure currents in the positive paths, so they will see currents that look matched (to within E2EUNB), even if the positive pair currents are unmatched.

If this is true, then the unmatched classes will confuse the PSE.

In this case, a 2p midspan in series with a 2p endpoint would work correctly and could take advantage of the mixed classes - but it would still have to manage Vdiff somehow.


> On Oct 10, 2015, at 4:20 AM, Yseboodt, Lennart <lennart.yseboodt@xxxxxxxxxxx> wrote:
> Hi Yair,
> We currently do not require the PSE to measure on the bottom rail. It has the choice to look at top or bottom.
> Therefore, as soon as a load shares either or both rails, it will depend on the PSE implementation  what current is seen. Not desirable.
> Dual load can only refer to loads that are separated on both rails.
> Bottom line: if a PD requests different classes, it needs to be fully (both rails) separated to be guaranteed to work.
> Kind regards,
> Lennart
> From: Yair Darshan <YDarshan@xxxxxxxxxxxxx>
> Sent: Saturday, October 10, 2015 12:46
> To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
> Subject: Re: [802.3_4PPOE] Analysis of E2EP2PRUNB with Dual Signature PDs
> Hi Chritian,
> Please my response below.
> Yair
> From: Christian BEIA [mailto:christian.beia@xxxxxx] 
> Sent: Saturday, October 10, 2015 1:00 PM
> To: Yair Darshan; STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
> Subject: RE: Analysis of E2EP2PRUNB with Dual Signature PDs
> Hi Yair,
> Referring to the picture on page 7, if the ground of the loads is the same, there is current sharing between the return lines (in blue).
> Yair: Correct but only on the return path as I showed. 
> So even if the  loads are different, the current on the return path is balanced, and may be higher then expected on the pairset with lower class.
> Yair: I guess you meant that the current is unbalanced etc. That is why this it has to meet Icon-2P_unb as I mentioned in slide 7 the blue text and also in the summary slides 8 and 9.
> So, if there is a common point between the pairs in the PD the two classes must be the same.
> Yair: I disagree. The class has to be the same only if there are two common points because only in this case there is a single load. If there is only one common path it is not a single load. Look that the positive path is isolated by the DC/DC components.
> The case were both positive and negative path has common path are described in Figure 1, Figure 2b and Figure 3b.
> Do you agree with that?
> Yair: Please see above.    
> Thanks
> Best regards
> Christian
> From: Yair Darshan [mailto:YDarshan@xxxxxxxxxxxxx] 
> Sent: Saturday, October 10, 2015 11:34 AM
> To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
> Subject: [802.3_4PPOE] Analysis of E2EP2PRUNB with Dual Signature PDs
> Hi all,
> I was asked to do this for Katania meeting.
> Please let me know if you have comments/questions etc.
> Regards
> Yair
> Darshan Yair
> Chief R&D Engineer
> Analog Mixed Signal Group
> Microsemi Corporation
> 1 Hanagar St., P.O. Box 7220
> Neve Ne'eman Industrial Zone
> Hod Hasharon 45421, Israel
> Tel:  +972-9-775-5100, EXT 210.
> Cell: +972-54-4893019
> Fax: +972-9-775-5111
> E-mail: <mailto:ydarshan@xxxxxxxxxxxxx>.