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

RE: [802.3af] Power isolation

I am not recommending to connect the screen directly to chassis. It is best
to connect it through a resistor with series capacitor.
The capacitor will block DC and prevent potential ground loops.
The resistor will serve as termination to reduce EMI etc.


-----Original Message-----
From: Wayne Howell [mailto:wayne@xxxxxxxxxxxxxxxxxxx]
Sent: Friday, May 03, 2002 10:49 PM
To: Scott_Burton@xxxxxxxxx; stds-802-3-pwrviamdi@xxxxxxxx
Subject: RE: [802.3af] Power isolation

Hi All,

I've recently been dealing with isolation issues for the PD end of 802.3af.
I am very unclear on whether the isolation requirement of 1500V also applies
to the screen of the cable if STP is used.

The convention in many products has been to connect screen to chassis (frame
ground). Depending upon the product, this seems to vary from a direct 0 ohm,
to around 100 ohms.

Any clarification on this would be appreciated.

Wayne Howell
Artistic Licence (UK) Ltd 

-----Original Message-----
From: Scott_Burton@xxxxxxxxx [mailto:Scott_Burton@xxxxxxxxx]
Sent: 30 April 2002 13:47
To: stds-802-3-pwrviamdi@xxxxxxxx
Subject: RE: [802.3af] Power isolation

Greetings All,
I was considering the isolation question recently, with respect to the next
draft of the standard, and was referring to Yair's email below.

D3.0 of our document has the accepted text regarding isolation in 33.4.1.
This indicates that 1500Vrms dielectric is required between the port device
circuits including frame ground, and all MDI leads. This is essentially the
MAU/PHY requirement that appears in various clauses throughout 802.3. The
clause that is missing from our D3.0 document is that regarding Environment
A and B isolation for repeaters, but I believe this is to be included in
the next draft.

There seems to me to be an inconsistency when looking at these two
requirements together. In my understanding any repeater will have a MAU or
PHY per port. If so, then why is an Environment A repeater permitted to
have only 500Vrms isolation from MDI to frame ground when the MAU/PHY
requirement is 1500Vrms from MDI to frame ground?

Any clarification of this issue would be much appreciated!

Yair Darshan <YairD@xxxxxxxxxxxxxx> on 02/23/2002 11:17:33 AM

To:   "'Wayne Howell'" <wayne@xxxxxxxxxxxxxxxxxxx>
cc:   stds-802-3-pwrviamdi@xxxxxxxx (bcc: Scott Burton/Kan/Mitel)
Subject:  RE: [802.3af] Power isolation

Hi Wayne,

I'll summarize the the technical status of the isolation issue and than
refer to draft 3 section 33.4.1.

1.    According to the current IEEE802.3 standard, the MDI leads should be
isolated from the port device circuits and frame ground. (in both sides of
the MDI).
2.    In environment A there is no requirement for isolation between port
port.In Environment B there is a requirement for isolation between port to
3.    Now with the addition of the Power Over MDI technology (IEEE802.3af),
the above applies too, however, it implies how to treat the PSE circuits
the PD circuits.
       In environment A, the PSE can use a single main power supply that
delivers power to all ports. The power supply output should be isolated
       frame ground and from port circuits (the circuits as defined by
3.1    Due to the fact that a single power supply output is connected to
ports, there is no port to port isolation and it is OK according to (2).
3.2    From the above info, it is required to have isolation between PSE
output port circuits and Port circuits.(they are not on the same ground!)

4.    At the PD side, if we read the wording of IEEE802.3, the MDI should
isolated from the DTE circuits and frame ground.
4.1  When adding a dc/dc converter that its input is connected to the MDI
and its output is connected to the DTE circuits, the DC/DC converter
       should have isolation between its input and output in order to meet

4.2    From safety standards point of view, if the PD is closed with
package (no frame ground or metal enclosure) and there is no metal  parts,
        accessible to the user, there is no need for using isolated DC/DC
converter (which cost more than non-isolated one), however not using
isolated DC/DC
        converter violates IEEE802.3 which requires isolation between MDI
DTE circuits.

According to the above, I believe that the wording used in draft 3, section
33.4.1 need to be modified accordingly.
Possible wording could be:

"The PSE output shall be isolated from port circuits, including frame
(if any).
Port device circuits, including frame ground (if any) will be isolated from
all MDI leads.
PD shall provide electrical isolation between port device circuits,
including frame ground (if any) and all MDI leads."  or equivalent wording.

See the following link for more info:




     Darshan Yair
Chief  Engineer
PowerDsine Ltd.  -  Powering Converged Networks
1 Hanagar St., P.O. Box 7220
Neve Ne'eman Industrial Zone
Hod Hasharon 45421, Israel
Tel:  +972-9-775-5100, Cell: +972-54-893019
Fax: +972-9-775-5111
E-mail: < mailto:yaird@xxxxxxxxxxxxxx <mailto:yaird@xxxxxxxxxxxxxx> >.
< <> >