RE: [802.3af] Diode bridge both on the spare and data pairs
You are correct.
We are discussing here two problems:
1. Back powering/Leakage
2. Polarity insensitive.
Being polarity insensitive i.e. using diode bridge in both spare and data
pairs solve back power/leakage in some PSE implementations not in all.
In order to solve leakage problem in all PSE implementations, it is required
to use single termination for each port and not tying few ports to a single
termination. We put a cautionary note for it in Dallas and it looks that
adding diode bridge fix the problem in some existing switches and solve the
polarity insensitive issue as well.
If you see problems in mandating polarity insensitive on both pairs at this
stage of the standard, it will be OK to add informative note to recommend
this implementation and not mandate it however mandate it is the best
technical approach in my opinion.
From: Mike_S_McCormack@xxxxxxxx [mailto:Mike_S_McCormack@xxxxxxxx]
Sent: Sunday, April 06, 2003 4:13 AM
Subject: RE: [802.3af] Diode bridge both on the spare and data pairs
Personally, if we are trying to limit leakage, we should limit leakage.
Polarity insensativity does not necessarily mandate limited leakage, maybe
one implementation of it does, but will every implementation? If you are
lobbying for polarity insensitity, lets discuss that on its own merits.
The problem that has been identified is leakage / back powering which is
quite seperate from polarity insensitity.