Hi All,
I have red the last correspondence
regarding safety aspects and would like clear few issues.
- IEEE802.3af
as part of IEEE802.3 address two environment conditions that affects the
isolation requirements of a PoE port and address the case of which part of
the system segments (e.g. cables) are located outside the building or
crosses buildings.
1.1 The requirements of these environments deal with
the isolation voltage and the test requirements.
1.2 Regarding other aspects of the safety issue
such as power limit, voltage limit and current limit it addressed in the
IEEE802.3af spec:
1.3.1
SELV – Voltage is limited to 60V max at operating voltage is limited to
57V.
1.3.2
Limited power source: Limited to <100VA as per EN60950 definitions. We have indicated
at the Vancouver
meeting that the max value
will be less or much less than this value.
1.3.3
The PoE port is equipped with current limit protection at 450mA in
addition to secondary protection as per the limited power source requirements. The
PoEp current at its worst case will be still less that 100V/44V, or 100V/51V
etc or less that 1.3A per conductor per UL spec which is even lower.
- All
other protection circuits needed to meet surge voltage or other immunity
test voltages are implementation specific and it is practiced per vendor
product definition which is out of scope for us.
- The bottom line: IEEE802.3af is
design to meet safety requirements. PoEp will meet too these requirements
as long as the topics above are met and maintained.
Yair
From:
owner-stds-802-3-poep@xxxxxxxx [mailto:owner-stds-802-3-poep@xxxxxxxx] On Behalf Of Michael McCormack
Sent: Wednesday, February 02, 2005
12:26 AM
To:
STDS-802-3-POEP@xxxxxxxxxxxxxxxxx
Subject: Re: [8023-POEP] Cabling
and wire current capacities
I am on the hook for supplying test data on the cable temperature rise of
cables in bundles caused by adding power. I will not be testing what the
cable rating should be.
Mike
Please respond to "IEEE 802.3 Power over Ethernet plus"<stds-802-3-poep@xxxxxxxx>
Sent by: owner-stds-802-3-poep@xxxxxxxx
To:
STDS-802-3-POEP@xxxxxxxxxxxxxxxxx
cc:
Subject: Re:
[8023-POEP] Cabling and wire current capacities
I think that
there is a more rational approach here. Rampant
conservatism is not the road to progress.
The point is to determine,
with reasonable judgement and consideration, the
actual ampacity based
on factual information whose underlying
assumptions are clearly
documented.
I know that
the NEC (National Electrical Code) considers this wiring to
be Class 2 (Inherently limited < 100VA)
in article 725. I cannot find
another NFPA (www.nfpa.org
<http://www.nfpa.org/> ) standard that gives
more detail about a case like this. Does
anyone know of one?
I remember
that someone was going to see if they could get actual test
data - and I am not sure if it was Alan Flatman or
not.
Regards,
Martin
Martin
Patoka
Systems
Engineer
Texas Instruments
214-567-5487
mpatoka@xxxxxx
-----Original
Message-----
From: owner-stds-802-3-poep@xxxxxxxx
[mailto:owner-stds-802-3-poep@xxxxxxxx] On Behalf
Of Steve Robbins
Sent: Monday, January 31, 2005 9:08 PM
To: STDS-802-3-POEP@xxxxxxxxxxxxxxxxx
Subject: Re: [8023-POEP] Cabling and wire current
capacities
Guys,
This is just
a suggestion, but instead of consulting NASA specs, I would
ask UL (Underwriters Lab) or some other group
whose opinion will count
when it comes to fire safety. (I spent many
years working on the
electric power system for the international space
station, and don't
have a lot of trust in specs written by those
"rocket scientists" at
NASA.)
Here's what
makes me a bit nervous about the ampacity discussions I've
been reading in these emails: One of the
great things about PoE is that
it's the first international power standard, but
one of the scary things
is that it's the first power standard where most
hook-ups will be done
by some Joe Blow IT guy instead of a licensed
electrician. There will
be people using cables that are not in the best of
condition, hanging
them over sharp metal edges, hanging them near
heat sources, tying them
into large bundles, doing bad crimp jobs, hooking
them into patch panels
with loose screws, etc. There will be plenty
of instances where
something is getting a lot hotter than you expect.
All it might take is
a few small fires scattered around the planet for
big insurance
companies to start putting clauses in their
policies for corporate
customers prohibiting the use of PoE in office
buildings. That would be
the end of this technology.
I'm sure
that SAE or MIL standards will be useful as guidelines. But
when it comes to safefy in electronics, people
look for the UL or CE
marks on products, not NASA, SAE or even the IEEE.
I don't know if UL
or CE have specs for wire ampacity, but I'd feel
better if I knew they
were part of the discussion.
Anyway,
that's my two cents. Sorry if it seems alarmist. I just think
we should be very conservative about ampacity.
More conservative than
organizations who might have effective veto power,
such as big insurance
companies.
Steve
Robbins
-----Original
Message-----
From: owner-stds-802-3-poep@xxxxxxxx
[mailto:owner-stds-802-3-poep@xxxxxxxx]On
Behalf Of Derek S. Koonce
Sent: Monday, January 31, 2005 5:20 PM
To: STDS-802-3-POEP@xxxxxxxxxxxxxxxxx
Subject: [8023-POEP] Cabling and wire current
capacities
As I stated
I would do, here is information I have on the
cabling and wire current capacities. I scanned in
some documents and,
fortunately, found the complete NASA related
document. I cannot attach
the files to this email because the server limits
to 100K of
attachments. However, those that would like the
files I will be glad to
forward upon request.
The key
document is the MIL-W-5088K. This document has been
transferred to SAE control under standard AS50881.
The SAE web site has
a cost of this document, non-member, of $59. Maybe
someone on this list
has this copy or can obtain it to share with the
list.
The NASA
TM102179 document discusses the cabling design for
space payloads. I feel this would be the best
starting point.
http://snebulos.mit.edu/projects/reference/International-Space-Station/T
M102179.pdf
Surprenant
is a cabling company that has some data on cabling
derating. But their ampacity chart starts at 18
AWG and goes up from
there.
A fusing
current document was passed to me, years ago, from a
line of engineers and is more of an eye-opener.
The fusing current for
26 AWG wire is 20.5 A. Quite a bit to look
at, but clearly unreasonable.
Looking at
the NASA document (referenced above) and running the
numbers for 100% wire usage, 50 F temp rise we would get a
maximum
current loading of 3.588 A per wire. (A 4.46V
drop based on 2-pair
conduction for 100 m)
This is very
large and would put a maximum limit on the power.
This does not take into account for voltage drop
along the cable.
A bit more
searching brought up the following website
http://www.powerstream.com/Wire_Size.htm
This states
that 26 AWG wire can handle up to 2.2
A for chassis
wiring and 0.361 A for power wiring. These currents relate
to a voltage
drop of 2.74 V and 0.45 V for 2-pair conduction
for 100 m.
If we look
at any baseline for current in the wire. I would say
the 0.361 would be a good start. Then add on
upping the voltage to 55 V,
we can see about 39 W on the PD side.
Derek Koonce
Architect,
Standard Product Group
JSI
Microelectronics
4235 Forcum Ave., Ste. 500
McClellan, CA
95652
916-648-2089
x114
P