Jim and All,
For what little it may be worth, I'll toss in a few comments "from the
field" on this topic:
1. I have had two different types of experience with CAT 5 cabling and
the associated 10Base and 100Base-T Ethernet signals. The first was here
at ye olde salt mines where I, along with several colleagues, have had to
measure and evaluate emissions from Ethernet interfaces for the past
several years, relative to aerospace standards. The other was at home,
where although my Internet access there is via satellite dish, the
resulting Ethernet for connection is routed via unshielded CAT 5e cabling
that I installed when the house was built in 2000.
2. Aerospace emissions requirements are generally to either MIL-STD-461E
(soon to be "F", as 461F was just released) requirement RE102 or RTCA
DO-160E, Section 20 (for commercial). There are varying levels of
allowable emissions in both documents, with variances due to type of
application, placement on a given platform, and (for MIL-STD-461) branch
of the military. However, all of these requirements are more stringent
than FCC Part 15 requires, in some cases being more than 40 dB more
stringent.
3. As most of you probably know, FCC Part 15, Class A devices are NOT
intended for home use. Unshielded CAT 5 with standard RJ-45 connectors
will usually meet Class A requirements, and with care in design, can also
meet Class B requirements (Class B being approved for use both at home or
business).
4. As was stated in another posting, shielded CAT 5 cables are NOT
connected to a chassis ground via one of the 8 conductors. The contact to
chassis ground is via the metal shrouds of the plug and mating shielded
RJ-45 socket. This configuration does not meet the requirement for a low
impedance connection at all frequencies of interest, but it can help
reduce emissions from the cable if there is a limited frequency range of
interest (such as one or a few ham bands).
5. Emissions from 100 Base-T are much more difficult to control than
those from 10 Base-T.
6. Believe it or not, there is a fair amount of aerospace and military
equipment on the market, or on the way to market, that has 100 Base-T
Ethernet interfaces. The fact is, though, that those interfaces do not
and can not use RJ-45 connectors. The usual connection is via either a
sub-D connector or aerospace circular connector, both of which must have
EMI backshells and very short, low impedance terminating means for the
cable shields. What I am saying here is that if your equipment meets FCC
Part 15, Class A and/or B requirements, and you are receiving RFI from
that equipment or the interconnecting wiring, then the emission level is
obviously greater than what you can tolerate. Jim has presented many ways
to reduce emissions on the cables. There is another alternative: greater
separation distance between your router or hub, the CAT 5 cabling, and
your antennas. Keep in mind that Part 15 measurements are made at a
distance of 3 meters for Class B and 10 meters for Class A. For those
having indoor antennas or antennas located adjacent or over their houses,
the separation distance is not enough to prevent RFI to your station.
7. At home, I am fortunate enough to have all my HF antennas, except a
10m vertical, not closer than 50 or 60 feet to the house. (That vertical
is at the far end of the house and is at least 20 feet away from the
nearest CAT 5 cabling.) Some of the CAT 5 passes through or near the
shack, but no problems. The only RFI issue I have been able to identify
and confirm as being due to the Ethernet signals is a birdie I hear on
145.39 MHz, which just happens to be the output frequency of our local
repeater. I do not hear that birdie on my base 2m radio (with its outside
antennas), but if I have my 2m HT on me and pass near either the hub,
satellite terminal, or any of the CAT 5 wiring, then the HT hears the
birdie. I have to be within about 3 to 5 feet of any of those items in
order for the birdie to be a problem.
Bottom line: If you can not get your antennas further away from the noise
source, then you should apply any of the various fixes that have been
discussed by Jim in which the problem is either eliminated or reduced. In
severe cases, you may have to go to the point of either disabling Ethernet
when you want to operate or perform "major surgery" by burying the hub,
router, or terminal in a fully shielded and grounded metal box, using
nothing but shielded CAT 5 for all cable runs, and replacing RJ-45
connectors with sub-D types having EMI backshells and decent shield
terminations. In can be done, but may not be easy. As usual, YMMV.
73, Dale
WA9ENA
"Jim Brown" <jim@audiosystemsgroup.com>
Sent by: rfi-bounces@contesting.com
01/12/2008 10:18 AM
To
"gd0tep@safe-mail.net" <gd0tep@safe-mail.net>, "rfi@contesting.com"
<rfi@contesting.com>
cc
Subject
Re: [RFI] Wired vs Wireless Routers for RFI
On Fri, 11 Jan 2008 16:27:45 -0000, gd0tep@safe-mail.net wrote:
>The screened CAT5 I use here is fitted with suitable plugs, complete with
>screen. The plug connects to the screen of the router or what ever it is
>you're plugging into. Most modern items using CAT5 sockets are fitted
with a
>screen connection.
In general, shielded ("screened" in British English) Ethernet cable is
standard in Europe, whereas unshielded CAT5 is standard in North America.
Thus, the Ethernet equipment and connectors you describe are the standard
in
Europe, but not in North America. Indeed, shielded CAT5 is not easy to
find,
and I've never seen an Ethernet connector with a shield connection. I'm
not
saying they don't exist, they're just not widely available to the general
public.
73,
Jim Brown K9YC
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
|