Letter From TR 41.4
Steve -
Below is a letter which I was asked to send to your task force from the
TIA TR41.4 VoIP Gateways and Infrastructure subcommittee. It is intended
to provide you and your group with some ideas from outside of the IEEE
community itself. Thanks for the consideration.
To: IEEE 802.3af Taskforce on Powering via
the MDI
From: TIA TR41.4 Subcommittee on VoIP Gateways and Infrastructures
In our sub-committee meeting on 17 May, 2000, TR41.4has directed that the
following communications be sent to you as advice and as a possible aid
in your task of determining the method and system for providing power to
IEEE 802.3 compliant devices via the MDI.
In the course of our discussions, we reviewed the status and state of the
work in progress within your taskforce. The consensus of the group
is that there are three items of advice we wish to provide to you.
First,that there are differing conditions under which powering might be
profitably be provided via either the “signal” pairs or the “idle” pairs.
These conditions, while including overlaps, also include cases which are
distinct from each other. We therefore would recommend that
powering methods be adopted so that an end device might receive power
over either set of pairs, dependent on the infrastructure
conditions. This implies that the source device need only supply
power over one set of pairs as dictated by the network infrastructure but
that the end device should be able to receive powering from either pair
set. This compromise worked well in the case of ISDN powering and
should serve the IEEE 802.3af community in good stead as well. It
also provides a mechanism whereby the current log jam might be
overcome.
Secondly, the discovery procedure should indeed be executed on the pairs
over which power is to be provided and that such discovery should ensure
that the end device is in fact a “powerable” device that is safe to
power.
Thirdly, it is very important that the discovery or powering procedure
should also ensure that the circuit may also be safely powered. By
this we mean that there is the potential, particularly in cable reuse
environments wherein the craftsperson may think that they are using a
point to point cable as defined by IEEE 802.3, but which is in fact not
such, and that it has parallel devices attached. A case in point
for this might be in an office complex wherein an ISDN system is being
replaced with an 802.3 system. The ISDN system might legitimately
have multiple S-Bus devices on the same 4-pair cable which terminates
using “RJ45”-type connectors. If that cable is reused for an 802.3
network, the endpoint might very well respond to the detection scheme
within the detection parameters but when power is applied, parallel
connected surge protectors left over from the ISDN usage might overheat
and cause a fire if subjected to the nominal voltage and if the current
is not limited correctly. This third issue is basically a safety
issue. Since in the telephony world, cable reuse is common, we view
this as a serious, potential issue.
We realize that this last issue in particular is more directly related to
IP Telephony, but its applicability is more general than this one
application.
Again, we thank you for your efforts and hope that the above comments may
be of some use and service in your efforts.
Sincerely,
Robert T. Bell
Chair, TR41.4