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

Re: [802.3_4PPOE] FW: Are diode bridges really needed (2).pptx

Diode bridges are not required and never were, but, in my opinion, the only people who would will be willing to remove auto polarity are the ones who have never worked with a product that did not provide it.  Find someone who supported a twisted pair product that did not contain auto MDI/MDIX before you seriously consider removing auto polarity for power.  Could anyone imagine the trouble we would be in if AC powered appliances were neither physically keyed nor polarity agnostic? 

I was one of the few who opposed Pat's comment, but I'll admit now it was a great idea.  If you are bothered about the energy spent in diodes you are free to use steering FETs instead or even relays; there is no requirement to use diodes in particular.  It is only the less altruistic among us that prefer cheap diodes to the more efficient solutions.



Michael McCormack

On Fri, Jan 31, 2014 at 10:13 AM, Chad Jones (cmjones) <cmjones@xxxxxxxxx> wrote:
Christian, the reflector should be able to handle attachments up to 2MB in size. 

All, I am forwarding this on behalf of Christian. He is looking for some help to decide if we can pursue removal of auto-polarity in the PD.

Christian, I am going to forward this to Pat Thaler (I’m not sure that she’s subscribed to our reflector). I will copy you. It was her comment in AF that eventually led to us mandating autoMDIX support.

Chad Jones
MGR, HW ENG, Cisco Systems
Chair, IEEE P802.3bt 4PPoE Task Force

From: Christian BEIA <christian.beia@xxxxxx>
Date: Friday, January 31, 2014 at 6:02 AM
To: Yair Darshan <YDarshan@xxxxxxxxxxxxx>, Fred Schindler <fred@xxxxxxxxxxxxxx>, Dave Dwelley <ddwelley@xxxxxxxxxx>, "Picard, Jean" <jean_picard@xxxxxx>, Chad Jones <cmjones@xxxxxxxxx>
Cc: "Yseboodt, Lennart (lennart.yseboodt@xxxxxxxxxxx)" <lennart.yseboodt@xxxxxxxxxxx>, "Abramson, David" <david.abramson@xxxxxx>, "Kousalya Balasubramanian (kousalya.balasubramanian@xxxxxxxxx)" <kousalya.balasubramanian@xxxxxxxxx>
Subject: Are diode bridges really needed (2).pptx

Hi Guys,

I’m trying to find out if having a defined voltage polarity at the PD PI is an achievable goal.

Please look at the short presentation in attachment. It summarizes the spec we have today.


After this brief research, my understanding is the following:


About Alt-A: reverse polarity at the PD happens only if a MDI-X cable is used and there is no auto-detect mode

About Alt-B: reverse polarity at the PD should never happen.


Are those assumption correct?


If they are correct, I would like to understand if auto-detect mode is a common feature or it is something “advanced”.


I can imagine that a Midspan probably won’t have an auto-detect feature. But what is the reason for using MDI-X cables between the midspan and the PD?


Thank you for any comment


Best regards




I’m not using the reflector since I’m not sure I can attach presentations there.

If it is ok I will forward this email to the reflector.




Michael McCormack
+1 978 394 7422