[RFI] 160M/Modem problem
svetanoff at earthlink.net
Wed Dec 19 23:25:09 EST 2012
Jim and All,
Yes, UTP is used almost universally now for high speed data because it
works so well. However, STP ("Shielded Twisted Pair") has been available
for quite some time, but in my experience, the extra cost of the added
shield is not worth it. Reason? The ubiquitous RJ-45 plug/jack combination
is a terrible way to treat a cable shield.
I can't go into details because of company proprietary issues, but I can
tell you emissions to/from a data cable are reduced when the cable is
shielded. This is especially true for 100 Base-T and above applications.
The "cure" for using the shield effectively is to get rid of the RJ-45
connectors (yes, even the "shielded" ones are not very useful) and go to
mil-style connectors with backshells and shield terminating rings. This is
because it is not possible to get extremely low impedance shield
terminations on RJ-style connectors. (Note that I said "impedance", not
"resistance".) Without a low-Z path to chassis ground (at each end), the
cable shields just remain as antennas to re-radiate.
The commercial world gets by (barely) with standard RJ-45 or "shielded"
RJ-45 connectors, but not so for MIL-spec requirements or in the vicinity
of radio communications. Since replacing RJs with MIL-spec is not an easy
task on most equipment, keeping data lines and antennas as far apart as
possible, plus use of the best quality UTP cable, will do as much good as
anything else that is practical to implement.
Dale Svetanoff, WA9ENA
Sr EMC Engineer
E-N-A Systems, LLC
Specializing in shielding applications, system grounding, and lightning
> [Original Message]
> From: Jim Brown <jim at audiosystemsgroup.com>
> To: Rfi List <rfi at contesting.com>
> Date: 12/19/2012 11:12:10
> Subject: Re: [RFI] 160M/Modem problem
> On 12/19/2012 8:17 AM, Missouri Guy wrote:
> >> CAT5, CAT6, and all variations thereof consists of four very good
> >> twisted pairs, each twisted at a different rate to minimize
> >> crosstalk
> >> from one pair to another.
> > Hi Jim, (This is off list because of a fork in the subject matter) :D)
> > I can understand the need/use of twisted pair. But,
> > are the pairs designed for different frequencies and/or different
> > impedences?....and that's why the different twist rates?
> Good question. CAT5/6/7 has a nominal Zo of 100 ohms, and is designed
> for wide bandwidth (so that it can transport digital pulses that make up
> the Ethernet data). The twist rate does NOT affect the Zo, but making
> the twist rates DIFFERENT minimizes crosstalk between pairs, and that is
> quite important in Ethernet. It is equally important if that same cable
> might be used to carry multiple telephone conversations, or DSL
> circuits, or audio signals.
> CAT5/6/7 is very good cable, and experience in the pro audio world have
> shown it to be as effective as shielded twisted pair in resisting the
> coupling of RF into problematic audio circuits.
> It should be remembered that since the earliest days of telephony, the
> sole means of rejecting crosstalk between the telephone system and the
> 60 Hz power system that ran parallel to it for hundreds of miles was the
> twisting of the telephone pair. Thanks to the long wavelength of noise
> source was 60 Hz, twisting was effectively accomplished by crossing over
> at alternate support poles. With the advent of broadcasting and more
> broadband noise sources, twisted pair cables in balanced circuits became
> the standard means of noise rejection in audio systems, and shielding
> was added for microphone circuits.
> In general, twisting is FAR more important (and effective) than the use
> of a cable shield in rejecting most forms of interference.
> 73, Jim K9YC
> RFI mailing list
> RFI at contesting.com
More information about the RFI