Andy,
There are two sources of radiated interference with rouiters: the power
supply (usually a switcher), and the actual wiring carrying the Ethernet
signals.
Eben though Ethernet uses a balanced line scheme (like the telephone
system), there are numerous opportunities for differential mode currents
to develop. (Slight variances in the twist of the cable pairs, impedance
mismatches at the connector interfaces, and common mode propagation of
clock frequencies from within the router that couple onto all interfaces
are all ways that radiated emissions develop in routers and similar
hardware.) When they do, radiation usually occurs. Aerospace
applications are very sensitive to Ether net radiated emissions,
especially now that several major aircraft are "fly-by-wire" and the
"wire" is Ethernet. From testing that I did several years ago, I can tell
you that 100 Base-T is a much worse (by 10 to 20 dB) radiator than 10
Base-T.
My house was bult in 2000, and in keeping with the times, I pre-wired it
with many runs of Cat. 5E cable and plenty of jacks around the house. (The
wireless routers were just coming onto the market and were rather pricey
at the time.) Although I do not have all of my lines active, I have had
minor problems with noise from my system at just one frequency: 145.39
MHz, which happens to be the output of the local repeater for which I am
trustee! My base radios don't hear the noise, but if I have my HT on my
person, it will pick up the digi noise if I walk near where the Ethernet
lines are run. I don't have any problems on HF, and I credit that with
the fact that all of my HF antennas are 50 to 250 feet away from the
house.
I can tell you that equipment running 100 Base-T Ethernet will NOT pass
aerospace requirements to MIL-STD-461E, RE102, by using unshielded twisted
pairs or RJ-45 connectors. (One problem with RJ-45s is that they do not
provide a truly low impedance means to terminate the cable shields.) The
best way to meet these requirements ( and to get rid of RFI from your
router or LAN) is to use shielded twisted pair CAT 5e cable and connectors
that offer a backshell termination for the shield. This would mean having
to hack out the RJ-45s and replace with sub-Ds or other connectors that
can offer better shield termination to chassis ground. It also means NO
plastic cases and, as noted by others, a well shielded and filtered power
supply, as well.
It's the old story of victim (your radios via their antennas) and source
(the routers via their wiring). You have to either increase the
separation distance between source and victim (which is what I have done
by placing my antennas well away from the house) or reduce the emitted
levels from the source (painful, but doable, if you don't mind some
serious work and possible mods to the equipment).
Always remember: when making for faster digital equipment, you are also
building a better transmitter, whether you want it or not.
73, Dale
WA9ENA
"Andy" <ingraham.ma.ultranet@rcn.com>
Sent by: rfi-bounces@contesting.com
04/26/2008 10:19 AM
Please respond to
Andy <ingraham.ma.ultranet@rcn.com>
To
<rfi@contesting.com>
cc
Subject
Re: [RFI] Router RFI
Why do routers cause RFI ...?
Wired routers do not intentionally radiate. But some do, probably because
they are poorly designed. Just like many other modern electronic devices.
Wireless routers are intentional radiators, so by definition they send out
RF, though they are supposed to be on frequencies that don't interfere
with
other services. Some do anyway, in part because of inadequate bandpass
filtering.
> My question, if I go with a completely wireless setup for 4 computers in
the
> house, will I have a good chance of eliminating the RFI??
All other things being equal (which they aren't), the wired router ought
to
have less RFI, and might be easier to improve by wrapping all cables on
toroids or through clamp-on filters. The wireless router must be allowed
to
radiate so fixing it might prove harder. There's no good reason to
believe
that a wireless router would eliminate RFI ... unless you know the
brand/model you get is RFI-free.
Andy
_______________________________________________
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
|